ML20151D083
| ML20151D083 | |
| Person / Time | |
|---|---|
| Site: | Three Mile Island  |
| Issue date: | 12/31/1987 |
| From: | Berger J TMI PUBLIC HEALTH FUND |
| To: | |
| Shared Package | |
| ML20151D074 | List: |
| References | |
| NUDOCS 8804130382 | |
| Download: ML20151D083 (165) | |
Text
{{#Wiki_filter:, ;
& - # ]
'!I 1,
/ ll i
,q i
A RADIATION MONITORING i t SYSTEM FOR NUCLEAR POWER PLANTS I I t (With Special Reference to the l I Three Mile Island Nuclear Power Station) '
. i.' .
1 Three Mile Island Public Health Fund December,1987
)
2dited by: Dr. Jonathan Berger Executive Secretary Three Mile Island Public Health Fund Scientific Adsisory Board , I Scientific Advisory Board { Dr. KZ. Morgan - Chairman Edward P. Radford, M.D. Dr. George Woodwell Dr. Thomas Cochran Dean Abrahamson, M.D., Ph.D. Ian McHarg ; John C. Cobb, M.D. g H. Jack Geiger, M.D.
.i
( l 88041303B2 880322 PDR ADOCK 05000299 P PDR r
l i l . A RADIATION MONITORING SYSTEM FOR NUCLEAR POWER PLANTS
; (With Special Reference to the Three Mile Island Nuclear Power Station)
Three Mile Island Public Health Fund December, 1987 Edited by: Dr. Jonathan Berger Executive Secretary ' Three Mile Island Public Health Fund Scientific Advisory Board Scientific Advisory Board , Dr. K.Z. Morgan-Chairman Edward P. Radford, M.D Dr. George Woodwell Dr. Thomas Cochran Dean Abrahamson, M.D., Ph.b. Ian McHarg John C. Cobb, M.D. H. Jack Geiger, M.D. i I s
o o TABLE OF CONTENTS - 4 PAGE PREFACE................ ....................................... 2 OVERVIEW AND FINDINGS.......................................... 5 Proposed In-Plant Monitoring Improvements.................. 6 Proposed Environmental Monitoring Improvements.............ll RecommendationsForMonitoringEndEffects................)l4 Recommendaticas For Improved Data Communication........... 15 I INTRODUCTION.................................................. 18 The Need For a Comprehensive Radiation Monitoring System......................................... 18 Operating History......................................... 19 CHAPTER 1: BASIC PRINCIPLES OF POWER PLANT EMISSIONS AND PATEWAYS OF EXPOSURE AND IMPACT................. 24 OVERVIEW...................................................... 25 * , Power Plant Emissions: Critical Nuclides.................. 25 Blochemical Nature and Radiotoxicity of Nuclides.......... 29 Pathways of Exposure...................................... 31 Other Environmental Impact................................ 33 , CHAPTER 2: IN-PLANT MONITORING................................ 34 Concerns of In-Plant Monitoring........................... 35 - Current Status of TMI-l In-Plant Monitoring Systems....... 35 Practical Problems With The Current Effluent Monitoring Systems............................... 37 Personnel Problems........................................ 48 Proposed Improvements for In-Plant Effluent Monitoring Systems........................................ 49 4 Improved Monitoring Procedures............................ 56 . Cost Estimate For Proposed In-Plant Effluent s Monitoring System......................................... 62 -
?' _
'i te e TABLE OF CONTENTS - Continu;d PAGE CHAPTER 3: ENVIRONMENTAL MONITORING.......................... 64 Purposes and Concerns of Environmental Monitoring......... 65 J Current Practice.......................................... 70 Practical Problems With Current Practice.................. 72 Proposed Improvements in Current Environmental Monitoring.................................. 78 Improvements In The Fixed Off-Site Monitoring Network......................................... 88 CHAPTER 4: REDUNDANCY AND BACK-UP RE QUALITY CONTROL.............................Q.UIREMENTS:
......................... 93 Back-Up and Quality Control Concerns...................... 94 Assessment of Actual or Potential Doses to Jroups and Sub-Populations............................. 95 Regulatory Compliance..................................... 97 Provision of Adequate Warning............................. 98
+ .
Quality Assurar.ce Program Cont rol Methods . . . . . . . . . . . . . . . . . 99 Quality Assurance Program................................ 101 Quality Control (0C)..................................... 108 Contribution of Quality Assurance and Quality Control to Redundancy, Back-up, and Safety............................................... 110 CEAPTER 5: MONITORING END EFFECTS........................... 111 4 The Need For End Effect Monitoring....................... 112 Types of Endpoints to be Monitored in Human and Animal Populations............................. 114 Specific Proposals....................................... 115 Genetic Effects on Plants................................ 120 CHAPTER 6: ROUTINE AND EMERGENCY DATA COMMUNICATIONS......... 122 Communications Concerns.................................. 123
-t o o f
i f TABLE OF CONTENTS - Continued PAGE Current Communications System ........................... 123 5 Problems with Current Data Communications................ 132 i I Emergency Operations..................................... 133 ; f Proposed Improvements In Data Communication.............. 135 : a CONCLUSION................................................... 141 ! REFERENCES CITED............................................... 43 , i-
- b. s s
l !' a i ,i 4 1 f 'I 1 , s -{ % f. T 1 A i e L l f 4 i . j-r 1 I 4 i i.
?
i ) _ _q .1 4 1 iii - m--. . ._ . . . - - - __ ._ . _ . .
e . Tables Title Page Table 1(a) Radioactive Inventories and Whole- After Page 29 Body Dose Conversion Factors (3200 MWt--PWR) Table 1(b) Summary of Critical Nuclides Page 30 Table 2 Cost Estimate For Proposed In-Plant Page 61
- Effluent Monitoring System For TMI-l (
Table 3 Summary Comparison of the Current Page 65 THI-l Environmental Monitoring Program and a Recommended Program Table 4 Coolant Detection Limits After Page 104 Table 5 Topic Categories For An Assessment Page 116 of the Frequency and Pattern of Cancer Events in the Area Around Three Mile Island Table 6 Radiological Information Needs For Page 128 Response To Emergencies At A Nuclear , Plant Site Figurer. Figure 1 Typical Analog-Based Radiation After Page 38 Monitoring Program , Figure 2 Proposed Effective Inplant Effluent After Page 48 Monitoring System
; Figure 3 Locations of Air Iodine and Air After Page 74, Particulate Monitoring, Around TMI, Before the Accident in 1979 Figure 4 Present (1985) Locations Of. Air After Page 74 Iodine and Air Particulate Moni-toring, Around TMI Figure 5 Locations for Monitoring of Air After Page 74 Iodine and Air Particulates In A Hypothetical Improved System Figure 6 Undetected I-131 Releases After Page 75 Figure 7 Locations of Thermoluminescent After Page 75 ,
Dosimeters (TLDs), Around TMI, _t Before The Accident In 1979
- iv -
J- J l I l Figure 8 Present (1985) Locations Of After Page 75 i Thermoluminescent Dosimeters (TLDs) Around TMI (20 km quadrant) Figure 9 Present (1985) Locations of Real- After Page 75 Time (SENTRI) Gamma Monitoring Systems Around TMI . Figure 10 Locations of TLDs in a Hypothetical After Page 75 Improved System Figure 11 External Gamma Doses After Page 75
.. I ,
Figure 12 Diablo Canyon Power Plant Emergency After Page 130 Assessment and_ Response System 4 ) a t M G p i
"b
-v-
PREFACE k Settlement of litigation arising out of the accidental destruction of the Three Mile Island Reactor #2 in 1979 resulted in the establishment of the Three Mile Island Public Health Fund. The Fund is under the supervision of United States District _ Judge Sylvia H. Rambo, in Harrisburg, PA. At its inception, the{und consisted of five million dollars and has been utilized, pursuant to Court authorization, for research into: (1) the specific effects of the TMI accident on the health of persons living in and around the TMI facility; (2) the long-term effects of low level ionizing radiation on human health; (3) planning for nuclear power plant emergencies in the future at .TMI; and (4) monitoring of. ionizing radiation released by the TMI facility reactors under normal operations and during emergencies. This report deals with item #4 monitoring. In 1983, the Fund contracted with the Academy of Natural Sciences of Philadelphia to provide a report on the subject of improved radiation monitoring. The Fund also retained The Institute For Energy and Environmental Research ( " I FF'J " ) of Heidlberg, West Germany to provide general consultation with respect to discrete questions involving radiation monitoring at nuclear power plants and at TMI. The Academy prepared a report and thirteen appendices entitled "Light Water Reactors Monitoring and Management In-Plant and Environmental for Three Mile Island." IFEU also supplied materials to the Fund. These reports and consultations on' subjects not covered in these
documents -are the basis for this report on radiation monitoring for nuclear power plants. Members of the Three Mile Island Public Health Fund's Scientific Advisory Board provided guidance during the preparation of this report. Dr. William Kirk of the USEPA TMI field station, and Dr. Bernd Kahn extensively reviewed this report. An advisory committee composed of a cross-section of persons living in and around the TMI f acility also provided extensive comment. Dr. Jonathan Berger, Executive Secretary to the Scientific Advisory Board of the Three Mile Island Public Health Fund, served as the editor of this report. The editor acknowledges the word processing skills of Juanita Thomas. The introduction and chapters one, four, and five are generic to all commercial nuclear power operations.'In contrast, , the other chapters of this report address the specific improve-ments for both boiling water (BWR) and pressurized water reactors (PWR) with specific reference to the TMI 1 PWR. All examples and suggestions for improvements refer to the current TMI 1 monitor-ing system. Changes in emphasis and priority from the source documents represent the conclusions of the Fund's Scientific . Advisory Board. Any substantive changes are referred to in the text. The principal investigators identified above are respon-sible for errors in fact and omission in the material prepared by them and their subcontractors. . i In presenting this report on monitoring, the Fund recog- _ l nizes that radiation monitoring by itself cannot protect the ,
. _q 1
_ .- l
public from radiation exposure; it can only reveal the acount of releases after the fact and trigger protective action or warnings to stay indoors or to evacuate. Public safety depends on both management and 'other personnel being competent and dedicated in the operation and monitoring of nuclear power plants to the protection of the public f rom any extra radiation, however lit-tie. The subjects of quality assurance and operating philosSphy are not addressed in this report. o 9 J ! l l l l
-4 4
l
. . . -. = - _ - . . _ , ..
t
. OVERVIEW AND FINDINGS The principal report of the Fund prepared by the Academy of Natural Sciences found that radiation monitoring systems of U.S. commercial nuclear reactors, although in conformance with ,
- i Nuclear Regulatory Commission requirements, could be substan-tially improved. The science advisors of the, Fund conciti'ded _
separately after extensive review that the NRC requirements themselves appear inadequate to provide for comprehensive routine and emergency radiation monitoring at U.S, commercial nuclear power plants. Based on the above the Fund recommends improvements in: coordination of in-plant and environmental monitoring; training and management of personnel; number, type, and deployment of monitor =; and communication and sharing of data + , i with the public. + l These recommendations are intended to enhance the cur- , l rent PWR (pressurized water reactor) monitoring systems and 1 i provide a comprehensive basis for effective radiation monitoring, ; at nuclear plants and TMI. The Fund defines an improved radia- - J ! tion monitoring system as one that will provide data obtained i from: in plant measurements; environmental samples; and surveys of human, plant, and animal end effects. Such a monitoring ! structure embodies sufficient back-up and overlap to police i recognized pathways of radiation exposure during normal plant j i operations and to maintain coverage in the event of a severe .; accident. d
-t Proposed In-Plant Monitoring Improvements At TMI-1 and other nuclear energy facilities, sampling problems and poor information coordination hinder accurate in-plant monitoring. The recommendations for improvement, with an estimated implementation cost of between seven and nine i million dollars, along with systems already in p' ace of (any plants follow.
- 1. Centralize effluent monitoring and control through I state-of-the-art digital computers that provide real time information at one console.
- 2. Use trained technical staff specialists to guide ,
j the monitoring program. l l l
- 3. Develop commitment of senior management to central-ize monitoring and shift the emphasis from that of meeting regulatory' limits to emission reduction through the use of monitoring data to guide reactor operations. -
l.
- 4. Use real-time gamma spectrometry for on-line effluent monitoring. (The report of the Academy concluded that this recommendation was optional at the present time. The science advisors of the Fund did not agree with this conclusion.)
o _ - = . . ._. -. .
- c. .
h 5. Provide continuous monitoring of air and liquid effluents including: - i i o Gross beta and gamma activity in air (activity per l unit volume, and total activity). Li i o Gross alpha activity in air (agtivity per u,' nit volume and total-activity). i o Noble gas activity in air (total activity released and distribution of noble gas radionuclides by 1 of each isotope), i o Radio' iodine activity in air (activity of I-131 and , total iodine Ci). i 1 j o H-3 and C-14 released in air (activity of each). c 1 o Information on organic vs. inorganic and on gaseous j vs. fraction of activity on particles. . s 1 j o Gross gamma activity (activity per unit volume) and ! spectral analysis in liquid (activity /radionuclide j unit volume, and total released). 1 o Grcss beta and alpha activity in liquid (activity / l ! radionuclide unit volume and total released), s i -t k .
?
4- O
- 6. Select gaseous effluent monitoring locations that are as close to the final release point as practical, but situ-ated before any area where effluents merge with high-flow non-radioactive streams. Isokinetic sampling systems should be designed with careful attention to detail, and to cover the required flow rate range. Grab samples should be obtained at each fixed monitor location to permit cross-chepking of mond' tor ..
and laboratory analysis results for identical samples.
- 7. Assure that the estimated detection sensitivity of a monitor for a given radionuclide is not an interference-free measurement but that it in fact.ored for the actual mixed radio-nuclide effluents to be monitored.
- 8. Sample iodine activity using charcoal-filters which can be removed for quantitative counting. Substitute silver
- enlite sampling media for charcoal because noble gas activity interferes with iodine counts. Assure proper filter identifica-tion, retrieval, and analysis in emergency conditions.
- 9. Collect grab samples for laboratory analysis of beta-emitting tritium, carbon-14, and strontium-90; for alpha analyses of plutonium-239 and neptunium-237; and for general quantitative radionuclide analysis by gamma spectrometry. When required, use special sampling assemblies such as gas bubblers or cold traps for tritium sampling, compressed-gas samplers for low t i
.*(
l -8 - I w w
- O activity sample. streams, and bubblers or Ascarite traps for carbon-14 sampling. -
- 10. Incorporate modern electronic technology into the gaseous monitoring system design. Such technology includes low noise integrated circuit signal processing, conversion from analog to digital signals at the detector st.ation, and [the installation of microprocessors at the detector station.
- 11. Use samplers with moving filters to track instan-taneous particulate releases.
- 12. Locate fixed liquid effluent monitors in each effluent line that discharges potentially radioactiva liquid from ,
the plant site. Monitoring will include all lines to the off-site environment from the waste treatment system, all sources in the containment and auxiliary buildings, and lines to and from any building in which radioactive materials are housed. Monitor continuous liquid release sources, such as secondary side processes and secondary coolant system clowdown and drain discharges. Monitor waste water frem floor drains, "hot" showers, building sumps, laundry waste, holding tanks, pool for fuel storage, and releases from the secondary loop.
- 13. For discharge lines containing liquids with radionuclide concentrations requiring dilution prior to release, 1
insure that the on-line monitor provides a signal for automatic J i
-9 -
l
I termination of effluent discharge if the radioactivity level exceeds a predetermined limit. - I l 14. Minimize sediment deposit collection,'and plateout i of solids in sample chambers of liquid effluent monitors. ! Provide for flushing and chemical cleaning of sample chambers without dismantling the system. For comparison of laboraticry
~
analysis with monitor readings, provide a means to remove a grab l
~
sample from the monitor locations. i
- 15. Devise a routine procedure to assure that grab :
samples are representative of the entire contents of the liquid batch to the discharged, and that resin beads in liquid effluent will be detected. . s
- 16. For emergencies, important considerations include l adequate instrument calibration, accuracy of measurements, and l appropriate personnel training. Calibration of effluent l monitoring equipment for analysis of post-accident conditions l Aside fecm elevated presents a special challenge. radiation ,
levels (100 Ci/L for noble gases and 1000 Ci/L for reactor coolant), the potential range of radiation energy levels also requires special consideration. Detectors must be able to provide a predictable response when exposed to high radiation fields; therefore, readout instruments should be located in areas that are not likely to be highly contaminated in the event of an accident. 6 t t Characteristics of post-accident conditions require plant and staff capabilities that go beyond routine monitoring requirements. Equipment such as radiation counters should be , available in protected locations for emergency use since routine monitoring equipment may be destroyed or badly contaminated. A ! major criticism of existing post-accident measuring systems is that, although the recorders may work well ,,during accid lent , conditions, human operators would be subject to unacceptable doses of radiation in conducting essential operations. Proposed Environmental Monitorina Improvements At an estimated implimentation cost of approximately 1.5 million dollars, recommendations for environmental monitoring improvements, at TMI-l and other facilities, to assure the , measurement of radioactivity escaping via currently unmonitored :, p ' pathways, i'nclude :
- 1. A substantial increase in the number of pressurized ion chambers used for real-time on-line monitoring, and the use of satellites rather than phone lines for the transmission of .
remote data to central locations. (The report of the Academy did not cover the use of satellite communications.)
- 2. Development of programs to monitor noble gases in the environment. ,
i
-t l
- 3. The determination of atmospheric monitor locations based on the air flow and physiography of the Susquehanna Valley.
- 4. Flow analyses of the Susquehanna River that take into consideration the braided stream pattern.
- 5. Monitoring for sediments behind dam,s and eel wiers.
- 6. Consider forest soil core samplings at various depths to determine if there is any build-up of long-lived radionuclides.
- 7. Develop a sampling program using indigenous organisms thatt grow in all seasons, have large surface-to-volume ,
ratios, bicaccumulate large amounts of radioactivity, and are important to humans. Encourage the U.S. ' Nuclear Regulatory Commission to enforce ceilings on the environmental accumulations of radioactive materials with long half lives, not their annual discharges.-
- 8. Sample a range of foods including: dairy foods, the water supply, fruit, vegetables, domestic meats, and poultry, wild game, eggs, garden and orchard products, fish, crayfish, and native wild fruits (berries, nuts, and mushrooms.).
- 9. Quadruple the number of TLDs (thermoluminescent dosimeters) currently placed in each of the sixteen sections ,
, .t around a plant. In settlements, place some detectors indoors.
Use direct readout dosimeters at TLD stations. Place TLDs with pressurized ion chambers; use some TLDs that measure beta dosage. (The reoort of the Academy recommended a doubling, rather than a quadrupAing of the number of TLDs.)
- 10. Insure the availability of portable air sampling units, pressurized ion chambers, and gamma spectrometers,
- 11. Develop a portable microcomputer system- so emergency survey teams can send real-time data to a central computer.
- 12. Emphasize flexibility, mobility, and real-time .,
feedback for emergency environmental monitoring.
- 13. Integrate environmental and in-plant monitoring with the centralized effluent monitoring system. Develop a console that displays the relationships between off-site and in-plant conditions. (The report of the Academy regarded this ,
recommendation as optional. The science advisors of the Fund believe it should be mandatory.)
- 14. Treat, all monitoring as emergency monitoring so that there is sufficient back-up capability to handle esents, even in the presence of instrument failures. (The report of the Academy concluded that this recommendation 'is unnecessary so long ,
..t l . .
l as there is sufficient back-up capability for all monitoring i systems. The science advisors of the Fund cid not agree.) ,
- 15. Consider plant shutdown if redundancy or system degradation reach levels where compliance, dose assessment, or warnings cannot be achieved or monitored.
i
. \' .
- 16. Maintain constant evaluation of the radioactivity monitoring program at every plant to insure that technology, training, and methodology remain state-of-the-art.
Recommendations For Monitoring End Effects In recommending the inclusion of end-effect monitoring, , the Advisory Board is well aware that the monitoring of humans L I and ecosystems to detect effects of exposures to ionizing ; radiation is not now included in nuclear power plant monitoring ' systems. Similarly, end-effect monitoring is not included in the environmental monitoring programs for other facilities from which carcinogenic or toxic materials can esecpe to the environment. . The Board strongly suggests, however, that such monitoring should be routinely implemented, not only at nuclear power plants, but at other facilities capable of releasing toxic materials to the l environment either t;u r i ng routine operation or accident l l l conditions. In making this recommendation, the Board does not suggest that such monitoring be under the direct control of the l operators of the facility, however the cost of end-effcct [ N
- 14 -
i ! monitoring, as other necessary components of the monitoring 4 i j system, be borne by the firm which conducts the operation and r 1 therby imposes unusual risk for its environsi The science ad- l visors of the Fun'd also note that the subject of end-effect moni-i toring was not cddressed by the report of thm Academy. ; k Currently, at TMI-l and most other. nuclear enk'rgy ,
- facilities, end-effect monitoring is not done. Because l end-effect monitoring is necessary for the accurate evaluation of the long-term effects of radirtion, the folJowing monitoring i i
activities are recommended: ! l
- 1. Obtain samples from and monitor selected individu-j als or animals regularly during life and after death. . .
i l*
- 2. Conduct continued surveillance of vital statistics .
l in nearby coummunities. Monitor for childhood cancers and ! 1 . ! abnormal pregnancy outcomes. ( i l
- 3. Monitor geretic effects of radionuclides on plants -
1 using the spiderwort P <descantia ) . ) i l Recommendations for Improved ' ] Data Communication ; i } , j No technical monitoring improvement alone will relieve { the current public mistrust of the nuclear power industry.- Public trust will develcp only when the industry officials can l demonstrate an efficient, safe, and effectively run monitoring - l i
l 1 process. A good way to accomplish this would be to permit.public l review of and participation in the complete monitoring process. Recommended components for including the public in the monitoring process include the followings j
- 1. Implement centralized in-plant and off-site effluent data collection and analysis. ,,
j'
- 2. Involve local citizens in planning a radiological monitoring program.
- 3. Involve local citizens in the collection of environmental data.
- 4. Establish a review group of local citizens, some with scientific credentials, to receive timely routine reports.
- 5. Develop easy-to-read public monitors.
- 6. Accelerate radiatien dosage reporting to the .
i utility, governments, and the public. '
- 7. Establish an independent monJtoring group to run the safety net of biological indicators. This group will act as a check on the other monitoring programs.
I f
- 8. Assist local of ficials,' reporters, and editors to j i provide improved routine releases to the public and improved !
1 j j information about nuclear plant accidents. ! l , f
't
.l t f
f
] ,
I
- t 1 ] I i '
3 i i 1 l , 1 i . d 1 l I 4 4 s l j 1
- - - - ~ - _ . - _ - . . . _ . _ - _ , _ . . _ . - _ _ . _
e # 4 1 k 0 t l INTRODUCTION i 4 1 l i 1 l a i l i a l 1 d I i 2 l 1 1 d ) 4 l
)
i I l l 1 i 1
- -b i
a i 4
,g k
i
The Need For A Comprehensive Radiation Monitoring System 4 The use of nuclear energy has brought both human benefit and misery: in war; in several commercial power plant accidents including that of Chernobyl but also the medical advances made by use of radioisotopes as tracers.
.. \.' .
The U.S. public has been openly skeptical concerning the willingness and/or ability of utility executives and public officials to effectively manage nuclear facilities to assure public safety or to be open and honest concerning vital matters of health and safety. Improved monitoring of management, staff, and radiation releases of commercial power plants is central to this public concern. The operating history of reactors has + . reinforced these concerns. Operating History Routine Operations Absolute quantities of radioactivity released during normal plant operations usually exceed calculated values because 1 of "unanticipated operational occurrences, such as operator errors and maintenance spills" and the wide variance in plant l clean-up systems. (Kahn, 1980). Rapid cycling of .ae reactor l 1 power level, or changes in the coolant system chemistry, can cause significant increases in the concentration of radioactivity of the reactor coolant. Such shocks can dislodge deposits of 6 radioactive residues in the pipes and inject them into the - -1 i l coolant; they can also greatly accelerate the process of cladding degradation. Consequently, the levels of radioactivity in the coolant and effluents of a given plant may vaty from day to day or week to week. Rates of emissions for all U.S. nuclear power plants have only decreased slightly if at all in the,past five years. Whalig, et al. (1967) reviewed releases of tritium and fission and activation products for a number of U.S. reactors and concluded (1) th>t NRC regulations have not produced a significant trend of emission reduction and (2) that the operating philosophy of each plant has'more to do with emissions than NRC regulations or emission control systems. Current evidence suggests that such low level emissions may cause more , human cancers than previously estimated. The current estimate is now nearly 1000 fatal cancers /million human RAD of exposure. (Preston and Pierce, 1987) Accidents The events at Windscale in 1957, TMI II in 1979, and Chernobyl in 1986 emphasize the necessity for a monitoring system that can detect routine low level emissions and severe accidental discharges, and provide the necessary data for determining the environmental and human health consequences of such releases.
.1
During the 1957 Windscale f i'r e , in-plant monitors did not initially detect the accident and the readings of an offsite monitor more than half a mile from the facility first alerted authorities to the severity of the accident (Eisenbud, 1973). The Windscale experience shows why overlapping and back-up s'jstems are required.
?,
At TMI II, high radiation fields in the plant caused the main vent stack monitor to go off-scale during the critical early period of the accident. The result was no direct measurements of airborne releases (Beyea, 1984). In addition, at TMI II off-site monitoring did not provide blanket dosimetric coverage, with the result that doubt remains as to the actual accident radiation releases (Beyea, 1984). , At Chernobyl, there was intense local contamination from the explosion, but as the plume was lofted rapidly rain-out and contamination also resulted hundreds and thousands of miles from the plant. This situation emphasizes the need for adequate evaluation of discharges at the plant and off-site on a continu- . ing basis. Injuries sustained by Chernobyl emergency response personnel indicate the need for adequate dosimeters on site. To determine monitoring improvements, this report examines how the current in-plant and environmental monitoring systems at TMI-1 track the movement of critical radionuclides. Technical recommendations concern the integration of in-plant and , Y l
. I
environmental monitoring, coverage of unmonitored pathways and end effects, and improved sampling methods. Recommendations for improved plant performance cover the senior management's commit-menc to emission reduction, centralized data control, technically competent personnel, and the timely and honest communication of monitoring data to the public. Implementation of the recommenda-tions for monitoring presented in this report will imp {ove . reactor performance, environmental protection, and public confidence. The principal objective of the recommended program is the protection of the public and the environment. The program has four objectives (Patrick and Palms, A1, 1987):
- 1. Assessment of the presence of radioactive materials or fields resulting from normal plant operations or accidents; assessment of potential or actual exposure doses and/or collective doses to critical groups and populations.
- 2. Compliance with authorized radioactivity limits and legal requirements.
- 3. Monitoring of plant operations and containment of releases, and the effectiveness of effluent control; issuance of a warning if unusual or unforeseen conditions are determined and, where s
. .t 22 -
.. o appropriate, initiate a special- environmental monitoring program. >
- 4. Expedient provision of _ credible information to local, state, and federal agencies and to the public.
i'. 6 4 e 9 J e 9 N
..=(
1
t T ', . t' m
- CHAPTER 1: BASIC PRINCIPLES OF POWER PLANT EMISSIONS AND PATHWAYS OF EXPOSURE AND IMPACT i
e l h 4 I l i 4 l t i i 1 l
. .- 6 24 -
, I j
. 0 I
i OVERVIEW l l Monitoring programs should follow the course of releases of radioactivity, or of direct radiation from a nuclear plant, in as close an approximation to real time as 15 possible. Monitor-ing both in the plant and in the environment is essential for rapid evaluation of and response to potentially harmful situations. Radioactivity inst rumentatic,n shouYd be up-to-date - and predictive in estimating the radiation risk to humans from both normal and accidental releases. The principal function of any monitoring system is to make as accurate as possible predictions of radionuclide distribution and ultimate human uptake (Patrick and Palms, A2, 1987). Power Plant Emissions: Critical Nuclides A nuclear reactor produces hundreds of radioactive nuclides including: noble gases, alkaline earth metals, rare earth elements, and halogens. The quantity of radionuclides present in a nuclear power plant depends on: the operating thermal power level of the reactor; fission yield; the extent to which the mix of nuclides approaches saturation in the reactor (i.e., how long the reactor has operated); the extent of fission pro. duct decay following shutdown; the rate and extent of fuel burn; leakage rate; and other factors. Fission yield depends on the nature of the recctor (e.g., ratio of U-235 to Pu-239, energy spectrum of neutrons,) but generally is calculated for each nuclide for a particular PWR or'BWR. Radionuclide production by b
a reactor is dependent upon the extent to which each nuclide approaches saturation (fuel change history), the total hours and power level of plant operation, and individual nuclide half-life. The leakage of vapors and gases (e.g., radioactive noble gases, tritiated water, and 14 C02) varies considerably among reactors. Because a reactor may operate at varied power levelsandhavemultipleshutdownperiodsandfuelchangecyc{es, estimation of the total nuclide inventory in the facility at any time is a complicated matter usually requiring sophisticated computc. algorithms, such as the Origen code (Bell, 1973). Radionuclides. produced in a nuclear reactor have half lives ranging from a microsecond to infinity. Those of less than a few minutes usually decay before they can be vented from the , stack and those with half-lives approaching infinity are stable isotopes. A few of the radionuclides produced by nuclear plants are also present in the atmosphere as fallout from nuclear weapon tests: Ru-103 (39.4 days), Cs-137 (30.0 years), and Ce-144 (284 days). Because the environment contains so many radionuclides, identification and measurement of those unique to nuclear power. plants is vital. Oftec a radionuclide source (whether another reactor or from fallout) can be identified by determining isotope ratios (e.g. Sr-89 (50.5d)/Sr-90 (28 y) or Cs-134(2.04 y)/Cs-137 (30.0y)] (Whalig et al. 1987). The most obvious of the power plant radionuclide emissions present in the atmosphere have a short half-life. The s
-t c ,
aset of decay is common to radionuclides produced by weapons testing. Probably the most important of _ these are the iodine radioistopes: I-131 (8d), I-132 (2.26h), I-133 (20.3h), I- 134 (52rr.), I-135 (6.68h), and I-129 (1.7 x 10 7y). Other important plar.t produced radionuclides include the noble gases, all of whirh (except Kr-85 (10.7y)} have half-lives shorter than a few days. Currently, weapons fallout contains very little S r',/-8 9 (50.5d), is essentially all strontium-90 (28y), and has a very s small strontium-89/90 ratio. A comparatively high strontium-89/90 ratio is characteristic of radionuclides from plant emissions or fallout from a plant accident such as Chernobyl
'Whalig, et al. 1987).
Radionuclides produced by nuclear plants and having ., somewhat longer half lives, which are virtually absent from weapons fallout, include: Sr-89 (51.7d); Cs-134 (2.04y); and corrosion products such as Cc-58 (71.3d), Co-60 (5.26y), Cr-51 (27.8d) and Mn-54 (303d). (Raciccesium remaining in fallout today (1987) is essentially all cesium-137.) Radioactive products in the reactor include those resultirig f;om irradiation . of the dissolved or suspended corrosion products from piping gaskets and other structural elements. The most important of these are radioisotopes of cobalt, chromium, iron, and manganese (Whalig, et al., (a) 1987). . The radionuclides most likely to be released from nuclear power plants, in either normal operation or accident con- , 27 -
h . , 1
)
ditions, are the noble gases (especially from boiling water l reactors) which are volatile and chemically inert, and tritium I (E-3 hydrogen) (Whalig, et al. (a) 1987). Tritium may escape either as volatile hydrogen gas or as water (HTO). Next most i likely to escape are some chemical forms of radiciodine, l particularly elemental iodine and organic iodine compounds, and j cesium. There is considerable controversy about the chemihal i forms of iodine that exist under reactor accident conditions. l l The types of cesium radionuclides which may be released include l Cs-137 and Cs-134 which are soluble in water and somewhat vola- , 1 tile; C-14 is released in gases and liquids but seldom measured. other radionuclides produced in reactors are not likely to be i released except perhaps in small quantities of non-volatile substances that are able to pass through filters and other , i purification systems. During accidents such as the one at Chernobyl many of the less volatile radionuclides such as Sr-89 and Sr-90, the transuranic elements (e.g. Pu-239 (24,400y), Pu-240(6,580y), Am-241(458y)], and uranium isotopes are released f in relatively small amounts (Whalig, et al., (a) 1987). However, l a few grams of a transuranic, like Am-241 has a dangerous alpha radiation dose. The total radionuclide inventory for a reactor is not as important a human risk factor as is the potential for release of a radionuclide from containment. The most important property in i radionuclide's potential for release is its volatility. Next most important is solubility; soluble radionuclides may be dis- s 1 e o charged as monitored liquid effluents. Non-volatile nuclides that are not completely retained by filters may be discharged as airborne or liquid effluents. Biochemical Nature and Radiotoxicity of Nuclides The hazard (radiotoxicity) of radioactive materihls taken internally largely depends on their biochemical nature. Some radioelements enter the body readily, others do not. Some radionuclides are eliminated rapidly, others are retained for long periods. For example H-3 has a biological half-life in the human body (time for the body to eliminate half of it) of 12 days while the biological half-life of Pu-239 is 200 years. Radiciodine is of particular concern because it is ~ . . absorbed into the body through the skin, lungs, or digestive system and it concentrates in the thyroid. Strontium radio-nuclides are concentrated in bone and have a t.iological half-life of 50 years in bone. Cesium passes into the body readily, but is distributed throughout the tissue mass, rather than concentrated . in a particular organ. Insoluble nuclides, particularly the alpha emitters uranium and plutonium, may be retained in the lungs for a long time. The long term risk of plutonium is to the bone and liver. The behavior of radionuclides in the environment, and l l the extent to which they are concentrated in ingested materials j
is also of great importance. The eating habits of various ethnic groups must be taken into consideration to when estimating the dose to a given segment of the population (Patrick and Palms, A2, 1987). sot.e groups such as the Amish, eat mostlp locally grown foods while others living in the same area obtain food from a wider variety of sources. 1 Although it is clear that certain radionuclides carry a greater potential risk than others, there is not complete agree-ment on which are the most significant. Tables 1(a) and (b) list a number of radionuclides and summarize the characteristic risk to humans carried by each. s G k
. 4D
_ . s
l Table 1( a-) Racloact;ve
- ventories anc Wnole-Bocy Dose Convers;on Factors (3200 MW:--pwa)
Shutcown Clo ud Irm a l.a tion , k t Cro m d g d uy.- Half-life Inwntery De/b ( r e s/C1-inn 418 4) 09 /$ Muc. lids a t (days) (106 cwi e s) ( w m3/C1-si (0-50 yrs) I rem-s2/C1) meeie C. ate 1 gr e 3.950. .M . 4 IV 43 .310f40 - gr.4ge . I F3 24 . 3 4 -01 .260( 4 0 gr a; .0529 47. .18 t( 40 . loot.41 _. g r-m .IIJ W. 46 M .00 .2 3 (.01 _ se-I 33 5.29 110. .90W-02 .100t 40 _ xel 25 .34 34 . M M -01 .1201 01 - ., ledines 4 . . .n 8.05 85. .872f.41 .600f 43 .10er.c3
!.I 32 . 0%d 120. .51I( .00 100f 42 .10 ;E .43 f.133 .3 75 1 10. . 3 54E.00 .20M 43 .311(.43 1 134 . 03M 190. .5II.00 . 300E .C .414 02 l-135 .250 150. 41W .00 .150E .03 .255(.43 Ceskas & heldlum c s-e as iw. 1.5 .50f 40 470f.45 .M96 44 C S-15 I3. 3.0 415 40 .59W .44 .41W .o4 Cs-131 11.000, 4.1 .122E .00 .35 4 .131(.04 an 46 18.7 .02% . 20 M -Cl . MW .04 .1854 43 I
fellvetum
& Ant many re-igi .M1 5.9 .93E-43 .34W.42 .s12 00 Tel2)u 1 09. .
1.1 .115 42 . 2 40E .04 .544E.02 Tel29 .048 31. 14 M -01 . 9EIK.cl .1 5 01 iel29s .340 5.3 .t u -G2 .300E.04 .24M 43 . ' f e-131a 1.25 13. .314 00 .550E.03 .960E.43 f e-132 3.25 120. 415f 41 150E.44 . 3M .44
$&.12 7 3.88 6.l .I51(.00 790(.03 .92M .03 56-129 .1 11 33. .265 00 .415 43 .1044 03 Alkallee fartftt 3r-es $2.1 94 - 410f.04 -
$r-90 11.030. 3. ?0 - .240f.06 -
Sr-91 4c3 110. .169(.00 .310f.43 .205(.43 84-140 12.4 164. 4444-41 .I90E.44 .3654 44 Volatile 0 ides (N)_ 6ew n. . 79 .216E.00 420fcS4 224f.44 Ce40 1920. .29 .600E .00 .820t 5 .5mt.44 me- 2.8 160. .364E-41 420E J .325l.43 fe e m .25 140. . 336f -0I . 93E .0I .I62f.42 . 4 103 M.5 110. .3111 00 .190E.04 . l l E .04 tv-105 .185 T2. . l ?9E .00 . Net.02 .794f.42 4 106 366. 25. .43t f 41 .620f.05 4%(.43 tr.-105 1.50 49. .l E2E-Cl .96McG2 .M M .42 henvolatile Osteet (td) r-w 4.ni 3.1 - .h.~488I - Y-91 59.0 120. .625f-43 .%et.04 .591(.01 le-95 65.2 150. .I621 40 . %0E .04 .IJ71.e4 fr-97 . 11 150. 4221-01 .520(.43 .5 M .43 10 -95 35. 150. .I6E.00 .I90E.44 .I64(.44 (4-140 1,67 160. .M M 40 .320f 43 .180(.44 Ce-141 32.3 150. .lG 41 .115 04 .lE2E.03 Ce-143 1. 3 1 33. .641( 4 1 . 340E .43 .224 43 C e-I 44 294 85. 431( 4 2 .320(.05 .I20t.43 Pr 143 13.1 130. - .820( .03 - 54 141 1I.1 64. .31 4 41 .19W.43 .30M 43 le-TM 2.35 1640. .300f 41 .250t.43 .202f 43 N43 32.500. .051 .525( 44 .13K.cs .620E .41 5 N4M J . 9( .06 .021 .230E 44 .820E.00 .253t.41 .- N-240 2.4(.06 .021 464f 44 . 82t .08 .5411 41 N441 5350. 3.4 411t 49 ,150(.07 .221E-42 8* 24I I.5(.05 .0011 46M -42 .860t.00 .1 G .03 W4 163. ,50 .500t-44
~
.190t .01 . 54E 41 C*4'4 M 30. .C23 .l 42T 42 4Et .46 . 34W.42 l l
Table 1(b) Summary of Critical Nuclides (Whalig, et al., (a), 1987) Physical Entrance Effective Biological Decay Into Primary Half Life Half Life Half Life Body Organ (Days) (Days) (Days) NUCLIDE Volatile Kr-85 readily none hours hours , 3,950 Xe-133 readily none hours _, hours I 5.28 ' HTO (H-3) readily whole body 12 12 4,500 I-131 readily thyroid 8 138 8.05 Non-volatile Cs-137 readily whole body 140 140 11,000 Cs-134 readily whole body 130 140 750 Sr-89 moderate bone 52 18,000 52.1 Sr-90 moderate bone 9000 18,000 11,030 Insoluble , Pu-239 low ,1ung (years)* 8.9x10 6 200 + ,
- depending on part of lung.
Pathways of Exposure Radiation can reach humans externally from direct radiation, or internally from nuclides which have been eaten, drunk, or breathed. Radionuclides released from a nuclear reactor facility, are usually found in higher concentrations near the l plant, and lower concentrations at greater distances. Direct Radiation Direct external radiation can occur either from airborne l radionuclides shortly after release (the radioactive
- cloud) from l
the nuclear facility, or from radionuclides deposited on the ground, vegetation, streets, houses, etc. Some radionuclides can enter the body by direct penetration of intact skin, includ-ing iodine and for HTO (Patrick and Palms, A2, 1987). Direct radiation from radionuclides in water (for example, to swimmers) is usually less important than airbhrne radiation because people spend only a small fraction of their time swimming and water is an effective radiation shield. On the other hand, radionuclides ingested in water can be a significant source of internal exposure. (Patrick and Palms, A2, 1987) Inhalation
+ .
Inhaled radionuclides can affect the lungs (especially if they are soluble), enter body fluids, and travel through the blood stream to other body organs. Transfer depends on the bio-chemical nature of the nuclide: insoluble particles may remain in the lungs, sometimes for extended periods of time. This is particluarly true of insoluble alpha- emitters, such as isotopes. of uranium or plutonium, (Patrick and Palms, A2, 1987). However, very small particles, less than about 1 micron, penetrate the ; 1 capillaries of the lung alveolae, and thus enter the blood stream and behave like soluable elements. , Large radioactive particles collect on the nasopharyn- . geal and tracheobronchial compartments of the lung and are later y _q 32 -
)
i swallowed, while the smaller particles (e.g., less than about 1 micron) pass directly to the lower lung compartment f rom which they may transfer to the lymph nodes and remain for a human lifetime. Ingestion Another cause of internal radiation is, ingested r ad'io- , nuclides. The primary sites of irradiation may be either the digestive tract, or organs reached through the blood stream. Ingestion may be via drinking water, milk, the surface of vegeta-bles, radioactive plants, or meat (Patrick and Palms, A2, 1987). Other Environmental Imoact The factors that cause damage to the environment also . . cause risk to man. Radionuclidos cannot harm the environment (and ultimately humans) unless pcwer plant containment structures and treatment and disposal systems fail. The radionuclides that are the most mobile in the environment have the most significant impact (Patrick and Palms, A2, 1987). Perhaps the most important factor in environmental impact is the tendency of some radionuclides 'to concentrate in specified portions of the environment. Concentration in flora or fauna is referred to as b.' concentration or bioaccumulation. Accumulation associated with an inanimate portion of the environment, such as mineral colloids, is referred to as sorption (Patrick and Palms, A2, 1987). s
._{
1 i e O e
.(
CHAPTER 2: IN-PLANT MONITORING l l I
% l
- b!
I l l
I Concerns of In-Plant Monitoring In-plant monitoring plays a key role in environmental monitoring for regulation compliance, providing warnings, and assessing population dose. The system must cover both designated and accidental release points, and must monitor each release point to determine instantaneous radioactive re, lease rates'(and _ cumulative release (Patrick and Palms, B.4, 1987). If in plant monitoring is defective or breaks down, determination of environmental contamination is greatly impaired. Under normal operation, information from .in plant effluent monitoring is obtained at release points. Radionuclides
~
in effluent samples are not mixed appreciably with rz.Jioactivity , arising from other sources (e.g., other facility releases or a weapon testing fallout.) In-plant monitoring makes it possible to determine whether effluents must be contained because of high radioactivity levels, or if another corrective action should be taken. Cumulative effluent monitoring records can provide an excellent basis for evaluating plant operations (Patrick and . Palms, B.4, 1987). Current Status of TMI-l In-Plant Monitoring Systems The fixed effluent monitoring system for liquid and gaseous effluent monitors at TMI-l is typical of direct-read (analog) systems which were installed in U.S. nuclear power s
-d 35 -
plants in the early 1970's. Thus, the TMI-l system has the inherent simplicity and standardization of such systems, as well as limitations in the areas of data manipulation, analysis, and storage (Whalig, et al. 1987). TMI-1, like most U.S. plants, has one liquid effluent release point and three to six gaseous effluent release points. The Radiological Effluent Technical Specification (a part of the legal requirements of the operating license) identifies all release points, and requires that each be monitored and the results reported to the Nuclear Regulatory Commission. The in plant liquid effluent monitoring program at TMI-l involves monitoring techniques that range from manual sampling and analysis, to continuous in-line monitoring with automatic , shutoff of liquid discharge if preset limits are exceeded. The principal atmospheric radiation monitors cover all gaseous effluent pathways except for the turbine building ventilation air, which in standard industry practice is not monitored or sampled. Atmospheric monitors vary between high range samplers, routine atmospheric monitors (with automatic interlock control , systems when noble gases exceed alarm set point), and an intake monitor intended to isolate the control room in the presence of high external radiation levels (Whalig et al. (b), 1987). An advantage of the TMI-l monitoring system design found in a majority of plants is the deployment of monitors with alarms , and termination functions on all tributary pathways to the plant ,
.t vent stack (fuel handling building air, auxiliary building air, and waste gas decay tank releases). This,use of monitors prior to mixing and final release offers more flexibility for control of releases, which is' absent when monitors are present only at the final release points (Whalig et al. (b), 1987).
The TMI-1 fixed effluent monitoring system can Neet current NRC-required effluent measurements if properly operated and maintained at all times. The system requires a high level of involvement and dedication by plant personnel to satisfy the needs of a rigorous effluent monitoring program. An inspection showed that the system appears to be in reasonable operating condition for a system of that age. But because many of the 15-year old electronic components of the system are becoming , obsolete, it will become increasingly more difficult and n expensive to maintain (Patrick and Palms, B4, 1987). Practical Problems With The Current
~
Effluent Monitoring Systems Problems associated with TMI-l and similar in plant effluent monitoring systems include the lack of rapid and accurate communication of data: sampling and measurement inadequacies associated with monitor techniques and design; and inadequate training and availability of personnel. The first two problems are linked to the third, because the ability of plant personnel to overcome problem situations will affect the j reliability of the monitoring system. 37 -
Data Communication Problems 4 Figure 1 shows the major elements of the "analog-based" monitoring system still in use at many nuclear power ~ plants, including TMI-1. Newer plants usually incorporate portions of the digital, computer-based approach proposed in this study. Fixed radiation monitors installed in gaseous and liquid effluent streams drive display meters with alarm setpoints in the control room, where other information on plant conditions and meteo-rological data are also displayed. Effluent grab samples usually are collected and analyzed by members of the radiological safety / radiochemistry staff, who use laboratory instruments and a small computer to calculate release concentrations and projected doses. Release information on gaseous and liquid effluents , usually is recorded on the laboratory computer disk and release permits and reports are sent to a document control group. The solid effluent management program is usually separately main- l tained and located in the solid radwaste processing area. (Patrick and Palms, B4, 1987) Communication of information required to generate release permits and monitor discharges usually is accomplished i either by telephone, or by visiting various locations. This l verbal transmission of data and recording of numbers on bits of l 1 paper presents occasion for miscommunication, and for the entry of incorrect data. (Patrick and Palms, B4, 1987)
.i 38 -
/
4
,, a. -# -#
OR8 OM - oR8 o t Id 8 - ORS o trd 8 - s- es ss es REACTOR CONTROL ROOM METEOROLOGICAL TONER , I r- 9 GASEOUS EFFLUENTS MON 110RS y _, LlOUlD EFFLUENTS MONITORS i * ., GASEOUS EFFLUENTS LlOUID EFFLUENTS
, TREATMENT SYSTEM TREATMENT SYSTEMS SOLID RADWASTE -
GAMMA LOW PROCESSING SPEC BKG SYSTEM BETA (HAND HELD INST.)
"^
COM$u'TER ^YEC S Figure 1: Typical Analog-Based Radiation Monitoring Program RADIO CHEM / COUNTING LAB From Patrick and Palms, B4,1987 s
. -t ,
1 l l
l R Maintenance and calibration of monitors is performed by an instrument and controls (I&C) group, and results are filed 1n the I&C office. Yet another functional unit, the radwaste treat-ment system group, operates the effluent treatment processing system, and maintains records of the amount, type, and processing history of radwaste in the system prior to release. With this f ragmentary approach to the collection, use, and storage of effluent release information, it is difficult for any single entity to gain an ove'rview or to develop a sense of responsibility for effective effluent control and monitoring. Even if each separate function were performed in a conscientious manner, it is unlikely that the separate efforts would mesh int'o an efficient effluents management program. . Sampling And Measurement Problems 1 Liquid Effluent Monitoring Grab Samples -- The radionuclide content of waste water tanks is measured by taking a "representative" sample of approxi-mately one liter from the tank. Because the tank size is in the thousands or tens of thousands of gallons, and suspended radio-active solids may form sludge in the tanks, it is extremely difficult to obtain a truly representative sample. Although
- 1. (This discussion is based on Section F.5.4 "Practical Problems With Effluent Monitoring Systems" from Appendix F InPlant Monitoring Systems For Radioactive Effluent For Nuclear Power Plants, in Light Water Reactors, Monitoring and Management for Three Mile Island Academy of Natural Sciences, -
o Philadelphia, 19.87).
~
standardized methods for trixing tank contents and prescribed recirculation times are used, slight variations in sampling technique can affect the radionuclide content of a sample. Also, 4 changes in the acidiuy of waste water due to operator error can cause the release of large slegs of nuclides previously built-up on the tank walls. Grab campling may miss the discharges, especially discharges of liquid containing radionuclides $ hat emit primarily alpha or eca radiation-Pure Beta and Alpha Emitters -- Liquid effluent monitors in the U.S. use a sample chamber with a reentry metal well into which a sodium iodide scintillation detector is inserted. Pure beta and alpha particle radiation and weak gamma radiation cannot be detected by the liquid monitor. Beta measurements are tiaken ., sporadically from the grab sample, but sampling or valve alignment error can result in the release of unmonitored pure alpha and beta particles. Because current units cannot detect all radionuclides, the monitor cannot provide a true real-time check againe: regulatory limits. Better technology is needed. As soon as reliable alpha and beta particle monitors are . available they should be installed. This is important for monitoring beta ermitters like C-14, H-3, Pu-241, and alpha emitters like U-235, Pa-239, Am-241, and other actinides. l Resin Beads -- it is not uncommon for resins from demineralizers to infiltrate the liquid waste system. A single contaminated resin bead, if discharged into the environment, s:
.6
- 40 -
could conceivably result in measurable radioactivity in humans. A small quantity of radioactive resin bende in a Waste tank would probably not be picked up in the grab sample under the present system, and would appea'r only as a momentary spike on the liquid effluent monitor. Background Buildup in the Monitor -- The interior wSils of the liquid effluent monitor chamber collect a sticky sludge residue when shower or laundry drain waste is released. Particulate matter also becomes trapped in the monitor chamber, particularly if careful attention to the flow pattern is not incorporated into the initial design. This buildup of radioactive material in the chamber increases the monitor background, thus reducing the monitor sensitivity for detecting , radionuclides in effluent. Radionuclide Soecific Measurement Requirements -- Although calibration using cobalt.-60 or cesium-137 is used to produce "m.icrocurie per ml" calibration curves, monitors cannot demonstrate compliance with the 10CFR Part 20 release limits . required by licensing. Existing monitors are not able to measure radionuclides with low energy or small emissions.- Since monitors detect gross gamma activity, high gamma ray counts can completely mask activity from radionuclides with low gamma-ray counts. This permits the undetected presence of low gamma radioactivity emissions in excess of the permissible concentration limit. Thus H-3, Pu-23), or C-14 releases could pass undetected and become a .
..t
- 41 -
significant environmental health problem. Improved technology is necessary to resolve this problem. - 1 Gaceous and Particulate Monitoring Anisokinetic Sampling -- Although most systems for gaseous effluent sampling claim to be isokine. tic (capable '( of equal flow measurement), a careful analysis demonstrates that few , realize this objective. Changes in daily a rtd emergency conditions make the effective design of isokinetic sample collection systems for particles, including particle-bound iodine, extremely difficult. Self-adjusting sampling systems are . accurate only within certain limits. Loss of Particles in Sample Lines -- The loss of particles from aerosols in pipes and tubing is well documented. The loss fraction is a complex function of the particle size distribution, pipe size, pipe surface, flow velocity, bends in the pipe, flow rate, temperature, humidity, and other factors. Exact losses are difficult to predict and radiciodine particles . are subject to the same loss mechanisms as particulate radionu-clides such as cobalt-60 or cesium-137. These particles may also undergo chemical reactions with the line tubing walls. Longer sample lines usually mean greater losses. This could make ' readings falsely low. Resuspension of deposited particles would l l l 1
also tend to increase the backround le ve l', and also give false low readings. - Measurements using actual sample lines at two nuclear plants showed loss fractions of 21-38% in one case (Necrenbeen, 1985), and 50-99% in another (Christianson, 1984). These loss fractions are significant for quantitative radioactiv'i ty measurements of effluents; additional study is needed on this subject. Cumulative Resoonse: Filters -- Because ordinary 1 monitors use a fixed filter to collect a cumulative sample, readings reflect the total particulate release since the last filter change. Thus, if the monitor reading is not corrected by , some function of prior readings, the monitor does not reflect the release for a 'specified time period. The most rigorous correction requires knowledge of the half-life of the radioactive particles on the filter and th'e half-life cannot be known from a gross count monitor. Particle bound iodine is collected in the particle filter which precedes the iodine cartridge. T correc- - tion must be made for iodine observed on the particulate filter; this cannot be done by a gross count monitor in real-time. Charcoal filters also retain noble gases from the sample l stream. The use of silver zeolite cartridges minimizes noble gas retention, but is now considered.to be cost prohibitive. Lower s
~
i l
cost silver-based gel cartridges are being evaluated as an alternative. . To date, there are no very satisfactory methods for obtaining instantaraous concentration estimates from gross count cumulative particc. te monitors. Mechanically moved filter systems are available and they should be used. Fixed sys) ems should be replaced by mechanical change filter systems. However, changes in collection and analysis procedures can assure more accurate measurements of radioactive iodine in samples. Cumulative Resoonse: Cartridaes -- Like conventional particulate monitors, iodine monitors respond to cumulative, rather than instantaneous releases. The amount of iodine , retained on a charcoal or silver zeolite cartridge is a function of several parameters, not all of which are considered in many sampling situations. Changes in plant operating conditions can also alter the parameters affecting iodine release rates making the performance of iodine monitors under accident conditions even less relaible than in normal operations. Under ideal sampling. 1 conditions, iodine is collected on the very front face of the j i sample cartridge, thus producing a well defined geccetry for i counting. Calibration can be accomplished by use of a cartridge in which the traceable standard is loade'd into a similar front "plane" geometry. However, sampling conditions which distribute the iodine through the cartridge, or lack of care in positioning m
- 44 -
the cartridge of the calibration standard for counting, can lead to significant errors in quantitative iodine analysis. - Sensitivity Interference -- For monitors in which particules are counted as the sample is collected, the presence of noble gases at much higher levels of activit- 'a n cause interference. Polyethylene Marinelli beakers _,show a stri5ng "memory" for noble gases due to adsorption of gases wi*hin the polyethylene r,cructure. Background count rate buildup in gross counting chambers is caused by adsorption of noble gases into gaskets, 0-rings, or other organic seal materials within the chamber (Paciga, 1985). Gas calibration standards in glass vials with rubber septum tops change in activity over long storage' periods. , , The presence of lines or cavities containing noble gases, or of organic materials which retain noble gases, can be minimized by effective monitor design. However, the presence of a volume of noble gas-laden air between the detector and the filter surface is inevitable, so noble gas interference cannot be - completely removed unless a more elaborate system is employed. l 1 Iodine filters retain radionuclides which can interfere with the counting of the ,I-131 peak in gross monitors, and can present excensive activity for gross counting or for radionuclide l specific counting. Much of this interference is due to the , higher-energy gamma emissions of the other' iodine isotopes and 4 , -E 2, l
, 45 - !
the fact that their relative concentration to that of I-131 depends on the operating history of the reactor. Typical-background subtraction methods provide a reasonable correction only for low activity levels and simple mixes of radionuclides, and only if appropriate correction is made for the various short-lived radioisotopes of iodine in relation to past reactor operation. .. '[ The very short range of H-3 and C-14 beta particles in air means that gas monitor efficiency for these nuclides is much lower than for nuclides with more energetic electron emissions unless the gases are passed through a counting chamber or scintillator. A step change in reactor power, the onset of routine fuel element leakage, or damaged fuel, can drastically change the radionuclide mix in a gas monitor. This is similar to the problem that occurs with iodine monitors. The shift toward increased amounts of short-lived fission products can interfere with the ability to adequately monitor longer lived radionuclides . with more important dose consequences. At the onset of fuel leakage, the increase in rubidium-88 with 17.7 min:ite half-life can swamp the detector, causing all other particulate components of the gas to be ef f ectively unmonitored. Since mid-range ,and high-range gas monitors usually detect gamma rays, their response factors (counts / minute per uCi/ml) are more strongly affected. 2
-t Background Subtraction Techniques --
Some particle monitor designa use additional detectors to correct for counts in the non-particulate component of the sample. Alpha counters can be used to correct the gamma-ray background interference if essentially all of the background is due to radon and its' daughter products. The background activity, computed as some ; multiple of the background detector (s) count rate (s), j' is : subtracted from the counts recorded for the filter. This presents several potential problems in practice. For example, plants which are experiencing fuel cladding fai'ure do see transuranics in the waste systems: the auc. cic alpha subtraction system would incorrectly reduce the beta-gamma count in the event of a significant transuranic content in the aerosol. Low Pressure in Measurement Chamber -- The normal design for particulate / iodine / noble gas monitors is to use an air pump on the outlet to pull air in series through the three sampling stations. For certain flow conditions, and particularly for heavy loading of the particle and iodine filters, the gas chamber can operate at some degree of vacuum. Since calibration is based . on standard pressure, operation at any lower pressure introduces a small error in the monitor reading. An NRC bulletin on this topi,c was directed to all nuclear power plants, and has resulted in the correction of this problem at most plants (NRC,1982). 1 i Calibration by Non-recresentative Mixtures -- The noble , gas monitor typically views a complex mix of radionuclides which ,
._q
- 47 -
)
. t va-les with the decay time of the gas following the fission process.- Only three gas calibration standards (Xe-127, Xe-133, and Kr-85) are normally available from the National Bureau of Standards, and these are available only occasionally. Thus, it is difficult to simulate many conditions under which the detector system must operate. Practical problems with presenting a known volume of calibrated gas mixture to the detector under controi;1ed conditions further complicate the process of obtaining accurate calibrations of gas monitors.
Personnel Problems suppliers of radiation monitoring systems have expressed a strong concern concerning the qualifications of personnel , J responsible for operating effluent monitoring programs in nuclear ' power plants. The assignment of personnel without proper regard to qualifications, and the frequent turnover in personnel, can and does have a significant impact on the effectiveness of radiation monitoring programs. As monitoring programs around the world and at TMI-1 become more comprehensive, the need for . properly trained support personnel will become more acute. (Patrick and Palms, 34, 1987) To meet regulatory release limits and achieve improved operational and safety performance, owners and operators of systems like those in TMI-l should upgrade data communication, sampling methods and measurements, and personnel qualifications. j - 48 -
4 A l v h 1 MET 1040LOGOL m ia [ ,
. . p: : ,, q Emuent EMS CONSOLI .- .
u m0Ctss0R A A suas MINTERi uwe ,terrtnt D7un7 oy7py7 anas O m :ss tirtumsa m m w, u
- u ssoa me C0eum .
SVTWARE FOR
-mas
-JtELEASE PERWei uhK TO
$0U0 --
-RECWtDs DATALSE [I CENTM RICORDS usTt -cCswworos CouPum ways mm W PROCES$0A UhK TO
---- 'jj CORPORArt OmCE C3x ans Re a Ds mu avu -
,,,e d, ENGGDC u=R r0RESP 0860 i em F w C"" C "" 1
- -- z~a :
I am F "" I i l Figure 2: Proposed Effective Inplant Effluent Monitoring System uvio Pooctss % From Patrick and Palms,84,1987 - C0wNm -q 1 I
The next section addresses these issues (Patrick and Palms, 84, 1987). Proposed Improvements-for In-Plant Effluent Monitoring Systems A variety of improvements are necessary to upgrade the TMI-1 in-plant effluent monitoring system. Centralized effluent monitoring and control would improve monitoring for compliance. An increase in the number of monitors and protection of these instruments frem high temperature and radiation fields would enhance accident and routine monitoring. Changes in staffing and procedures would improve the accuracy and reliablity of monitor-ing for compliance, warning, and dose assessment. Communications Figure 2 shows, in simplified form, a recommended in-plant centralized effluent monitoring and control system. This compute r- based digital system integratos the operation and 4 communication functions of all areas of effluent monitoring and
~
release management, and permits data analysis, i ) The Effluent Monitoring System (EMS) would t,e a computer j sys tert that provides i.nformation collection, procesring, and storage. This system would receive continuous input from fixed liquid and gaseous effluent monitors and periodic result.s from i the solid effluent measurement program and the laboratory for ' s
-t radionuclide sample. analysis. Digital links to other on-site l
t .
e . computers would provide direct access to real-time meteorological data and to operational parameters for performing release calculations and preparing release permits (Patrick and Palms, B4, 1987). EMS software or systems could process and store fixed monitor information, and display or provide hard copy of cur' tent _ and archive information in tabular and graph form. The system simultaneously would display monitor readings and alarm functions in the reactor control rood. Multiple disk drives will provide independent recording of all effluent monitor and release data (Patrick and Palms, B4, 1987). Relaase permits could be processed at the EMS console ., location. With this system, all real-time information would be available at the EMS console, thus eliminating the need for manual etriesnl of information from remote sites. (Patrick and Palms, 84, 191!?) The ENS terminal would be linked to remote sites to . , permit direct t .ar.smission of data f rom the ef fluent database to central records or engineering support groups. To make all effluent monitor readings instantly available at the emercency, control center during an accident situation, there would be a direct digital link with the emergency response center computer. . The same procedure would provide the necessary link to off-site real-time on-line data, stored TLD data, and the -meteorologic U k l l I
data base (Patrick and Palms, B4, 1987). An effective system, discussed in chapter six, has been developed for the Diablo Canyon plant. ! The recommended EMS would be more powerful, reliable, and sophisticated than the present . simple analog systems and would be composed of mostly standard, mass-producedhardwarehnd software components. The reliability of these standard compo-nents is high; software and hardware upgrades can be obtained as necessary; and maintenance support is readily available. Software Software development efforts must be planned for the * , entire system, and based on an operating system with extensive scientific language capability. A data base should be selected 4 only after careful review of overall system needs. Code can-be developed only after softuare functional and design specifi-cations are completed. Software must be verified and validated prior to use. A set of test data, incorporating all features of - the system and independently verified, should be generated and used to test the system each time the code is modified. These procedures will reduce the risk of failure or incorrect results (Patrick and Palms, B4, 1987). i
=(
Staffing : Implementation of this recommended electronic data gathering, review, and processing would require the addition of data processing staff specialists. There would also be a reduction in the number of technicians needed. Both actions depend on a commitment to excellence on the, pa r t of seplor j management, and a Wall trained and motivated staf f dedicated to keeping emision.S as low as possible and to the practice of preventive mainte4ance. Cetecter Systems t Updated ef fluent monitoring requirements would include ,
',, detector systems for laboratory measurements including effluent '
grab 3 ample measurements; fixed detectors to continuously monitor effluent streams; and portable detectors to survey solid radioactive wastes. Fixed monitors would provide instantaneous c information on gross radioactivity in effluent streams (Patrick and Palms, 84, 1987) and continuous monitoring of releases for . o Gross beta and gamma activity in air (activity per unit volume and total activity released)
; o Gross alpha activity in air (activity per unit volume and total activity released) 1 i -
n
-t i i
o Noble gas activity in air (tota'l activity released j and distribution of noble gas radionuclides by % of each isotope) ! o Radiolodine activity in air (activity of I-131 and total radio-iodine activity) o Information should be provided on organic vs. in-organic -and gasous vs. f raction , of activity}' on particles [ o H-3 and C-14 released in air (activity of each) o Gross gamma activity (curies 1) and spectral i analysis in liquid (activity of radionuclidos/ unit volume and total activity released) 1 1 o Gross beta and alpha activity in liquid (activity per unit volume and total activity released) . 1 All of these are not' currently monitored at TMI-1. 1 ] Laboratory System > The recommended laboratory measurement system must in- - clude radiochemistry facilities, counting instrumentation, and separation procedures for determining measurements. 4 Measurement System Purpose _ ^ Gamma Spectrometer Radionuclide specific ! (germanium detector) analysis of gamma emitters 1 in mixed nuclide samples - Liquid Scintillation Tritium analysis N Spectremeter '4 1 I
Measurement.fystem Purpose Low-background bata Measurement for pure - . Counter beta emitters Alpha Spectrometer Measurement for (solid state) transuranics The gamma spectrometer would be an essential part of the effluent measurement program because it would provide a fast, ' i simultaneous measurement of radionuclide concentrations 16 a mixed-nuclide sample. A well-shielded, high efficiency germanium detector should be used for effluent sample analysis, and . the computerized analyses should be transmitted directly into the system. The system must be calibrated with NBS standards for all sample geometries and counting positions used for effluent sampling. - To provide the required sensitivity levels for effluent measurements, the low level beta counter (GM) must have a background count rate of one count per minute or less. Commercially available gas flow proportional counters, sealed chamber proportional counters, and sandwiched scintillation detector systems could provide the desired performance capabili-ties (Patrick and Palms, B4, 1987). Several commercially available 1.iquid scintillation counter systems and solid-state (silicon) alpha spectrometer systems can be used for effluent measurements.
-t I
Fixed Monitors Fixed monitors should be installed in each in-plant sffluent stream which might discharge to the environment through a release point. Even safety valves which only discharge i occasionally should be monitored. These are not monitored at TMI-1. The real-time gross activity monitors. recommended for optimum monitoring are the proven, commercially available detectors such as ion chambers, GM tubes, proportional counters, beta scintillators, gamma scintillators, and cadmium telluride I detectors which comply with ANSI, ANS, and IEEE standards for nuclear plant use (Patrick and Palms, B4, 1987). Selection of detectors for specific applications must be . based on a sound, practical engineering analysis of detector e , capabilities, and of the actual conditions at sampling locations. The selection of a detector based on the measurement sensitivity [ t under idealized conditions for a specific radionuclide may not [ produce the desired sensitivity in practice because of interfer-ence from other radionuclides, background buildup, or use of an
- unrepresentative sample.
l Calibration of fixed monitors is difficult and the methods and extrapolations which are frequently applied have been shown to be deficient. At TMI-l calibration methods should be developed which are specific for measurements to be made by each - i detector in the system. Calibration with representative , ,' 1
radioactive sources over the entire range for which performance is required would be essential. Calibration should be based on the use of NBS calibration sources, or on applicable ANSI, ANS, and IEEE standards (Patrick and Palms, B4, 1987). Real-Time Gamma Spectrometry For In-Line Monitoring Identification and measurement of gamma emitting radio- l i nuclides in the effluent stream during discharge offers more i l information than is provided by the gross activity det.ectors now l l used for effluent stream monitoring (Patrick and Palms, B4, 1987). Currently, real-time gamma spectrometer effluent monitor-ing systems cost $1-2 million per installed monitoring station. The Fund's advisors concluded, in contrast to the Academy, that . I l over the life of a power plant this expense would be justified. Improved Monitoring Procedures 2 l Gaseous Effluents Monitor locations in each of the effluent streams must be selected as close to the final release point as practical, but prior to effluent dilution in high-flow non-radioactive streams.
- 2. (This discussion is based on material from sections F.S.4 through F.7 of Appendix F "In-Plant Monitoring Systems For Radioactive Ef fluent For Nuclear Power Plant in Light Water Reactors Monitoring and Management In-Plant and Envir¢al For Three Mile Island, Academy of Natural Sciences, Philadel- s phia, 1987.) -t
__ _-_____.___________.___.___.____m_ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _
Isokinetic sampling systems must be based on careful attention to t design detail, and must consider the entire range of flow rates related for isokinetic sampling. Grab samples obtained at each fixed monitor location woul'd permit cross-checking of monitor and laboratory results for identical samples. , i Monitor designs should incorporate commercially avdil-able detectors- such as GM, ion-chamber, proportional counter, beta scintillation, and cadmium telluride detectors. The selection of a monitor which gives adequate sensitivity for practical response times must be based on analysis of detector , performance specifications, sample volume and geometry, inter-ference counts, and the required statistical accuracy of the . 3
+
results. These considerations would determine whether an in- . line, off-line, or on-line monitor is preferable for each appli-cation. Particular care must be exercised to assure that the problem of interference of a monitor for an given radionuclide is t considered when establishing the monitor's sensitivity. t Iodine samples should be obtained from charcoal filters - < which are removed for quantitative counting. Silver ::eolite sampling media must be substituted for charcoal because noble gas activity interferes with iodine measurement. Monitoring of the 4 sample medium during collection is frequently accomplished using ; a gamma scintillation detector, but such monitoring does not provide quantitative data on release rates. . - 57 -
i e Grab samples must be collected for laboratory analysis of beta-emitting tritium, carbon-14, and strontium-90; for alpha and beta analyses of the actinides. Those of importance are Pu-239, No-237 Pu-241, and Am-241. Special sampling assemblies would be required, such as gas bubblers or cold traps for tritium sampling, compressed-gas samplers for low activity sample streams,andbubblersorAscaritetrcpsforcarbon-14sampliny. Recent electronic technology should be incorporated into the gas monitoring system design. Such technology should include low noise integrated circuit signal processing, conversion from analog to digital signals, and the use of microprocessors at the
. detector station.
Particle Monitoring Particulate activity should be measured by collection of particles in a filter sample system, followed by counting the filters in the laboratory. As noted for iodine monitoring, continuous monitoring of the filter by a scintillation detector . may be performed, but the real-time results do not reflect instantaneous release rates. Samplers with moveable filters will a overcome this limitation. 1 a 4 j %
-l
{ 9
o . t Liquid Effluent Monitoring b Fixed liquid effluent monitors should be located in each release line discharging potentially radioactive liquid from the plant site. Monitored lines would include all lines to the i off-site environment from the waste treatment system, all sources in the containment and auxiliary buildings, and any building housing radioactive materials. Continuous liquid release sources, such as secondary side processes and secondary coolant l system blowdown, should be monitored if they could contain i radioactivity under other than routine operating conditions. Such is not the case of THI-1. For discharge lines from sources that routinely contain * , liquids with radionuclide concentrations whien require dilution prior to release, th'e in-line monitor must provide a signal for autcmatic termination of discharge if the measured radioactivity level should exceed a predetermined limit. Although the Radiological Effluent Technical Speci f ica t. ions for all plants require that the monitor terminate the discharge whenever the - limits of 10CFR20 are exceeded, this is usually not physically possible, )' Liquid effluent monitors for gamma-emitting radionu-4 clides should be modeled on sodium iodide detectors mounted in a reentry well of a sample chamber through which the liquid .) i effluent stream passes. The sample chamber must be designed to C 1
- l J l
i minimize sediment deposition and plateout of solids. Provisions f for flushing and chemical cleaning of the sample chamber without < dismantling the system should also be provided. For comparison of laboratory analyses with monitor readings, a means to remove grab samples from monitor locations ~should be provided. r For critical applications involving automatic termin,'a- , l tion of discharge, the use of two independent monitors in paral-lel is recommended. This would have operational advantages such I as the ability to valve-out one monitor for flushing 'and background reduction while discharges are proceeding through the second monitor. Existing regulations, which allow for discharges ! even if required monitors are not available, should be made more restrictive to encourage plant staff to keep at least one monitor , in operation at all times. Grab samples should be obtained from effluent batches that are to be discharged. Laboratory methods must be estab- i lished for radionuclide analysis of grab samples by gamma spectroscopy, and for analyses for tritium, specific beta - , l emitters, and alpha emitters (e.g., Pu-239 and Np-237). A l l procedure should be devised to reveal whether or not the grab samples are ' representative of the entire contents of the liquid l batch to be discharged, and to assure that resin beads in liquid ; effluent would be detected, s r si
-b l
i l
j f Emergency Conditions 4 All of the above recommended changes in monitoring procedure apply primarily to normal operations. Serious acci- ; dents, however, would require important procedural additions. Important considerations would include adequate calibration, representative measurements, and adequate personnel t rain k'ng . , Calibrating effluent monitoring equipment for post-accident t conditions would be special challenge. Aside from elevated radiation levels involved (100 Ci/L for noble gases and 1000 Ci/L for reactor coolant), the wide range of radiation energy levels , (MeV/ dis) also would require special consideration. Detectors : must be able to provide a predictable response when exposed to f these high radiation fields (Patrick and Palms, 84, 1987). - ,
- Post-accident conditions would require plant and staff capabilities that go beyond routine monitoring. Equipment such J as radiation counters should be available in protected locations f for emergency use, since routine monitoring equipment may be destroyed. A major criticism of the existing post-accident -
measuring system at TMI-l and other plants is that, although the f recorders may work well during accident conditions, the human I operators would be subjected to totally unacceptable radiation doses (Patrick and Palms, B4, 1987). l 7 f. i i
I Cost Estimate For Proposed In-Plant Effluent Monitoring System . (Costs based on estimates by Whalig et al., (b) 1987) An estimate of the cost for a typical system of the type proposed is given in Table 2. Costs for upgrading an existing monitoring system in a nuclear plant would be somewhat less, depending on the amount of existing equipment thatcouldbeuked. . Table 2 Cost Estimate For Proposed In-Plant Effluent Monitoring System For TMI-l (Patrick and Palms, 1987 with additions) Item Estitaated Cost (thousands of Dollars) ,
- 1. Technical Specification Develop- 300 ment and Engineering Design
- 2. System Hardware 4,300 Gaseous effluent monitors 1,700 ,
Liquid effluent monitors 300 Solid waste measurement system 300 Laboratory measurement system 500 EMS computer & network system 1,500 Gamma spectrometer 2,000
- 3. Software 1,550 Software specifications dev. 100 Systems software 150 Applications software 850 V&V, test data set development 200 Interface to met, process comp. 200 .
Interface to records, corp. 50
- 4. Other 900 System documentation 200 System startup, calibration, and acceptance testing 500 Training program development 200 SYSTEM TOTAL ABOUT S9,000,000 v
..(
The proposed in-plant monitoring system'is designed to serve for both normal operations and emergencies with the underlying philosophy that emergency and routine monitoring should not be viewed separately and should be addressed in a unified manner. This includes the relationship between in-plant and off-site environmental monitoring. Experience has shown that ! in some cases the environmental monitoring syst,em provides $he ; first warning of plant emergencies. While the recommended i in-plant system would be the most concentrated and most timely way to control releases the recommended environmental monitoring system would provide invaluable data for emergency response, i long-term accumulation problems, and post-accident and normal operation dose estimation. l I t I e b
- I y I J .
j I l, P l 1 ,l I -9 i ) ^
O CHAPTER 3: ENVIRONMENTAL MONITORING '[ l .
4 O Purposes and Concerns of Environmental Monitoring Environmental monitoring provides data for population dose assessment, regulatory compliance, warning of unforeseen events, and public communication. The environmental monitoring network should have a real-time on-line and cumulative release monitoring capability. I' Currently used off-site monitoring programs vary greatly among nuclear facilities and countries. There is little evidence te suggest that facility operators follow any common theme in terms of specific nuclides, instruments, methods, media col-lected, or important population groups. The NRC stresses that minimum acceptable standards 'should be developed instead of , detailed regulations. This would permit the flexibility necessary to tailor systems to specific plant and environmental conditions at each site. Depending on the attitude of management and training and motivation of the staff it would result in innovation at some utilities and minimum compliance at others. The proposed improvements to the TMI-l environmental monitoring program are summari::ed in Table 3. e
- 6s -
Table 3 Summary Comparison of the Current TMI-1 Environmental Monitoring Program and A Recommended Program * . ATMOSPHERIC MONITORING Present TMI-l Program Recommended Program
- 1. Radiolodines 8 locations 9 locations and one mobile unit at present ,
- 2. Pa"ticulates: 8 locations at 9 locations monitoring for all present monitor for gross beta; those at left plus: carbon-14 gross alpha; St-89 and Sr-90 and actinides
- 3. PrecipitLtion: 5 locations at 5 locations as at present moni-present monitoring for: gross toring for all those at lett beta; gross alpha; H-3 (tritium); plus: I-131, and actinides Sr-89; and St-90
- 4. Krypton-85: 5 locations 9 locations and a mobile unit at present
- 5. Direct Radiation: Translumi- TLDs increased to 150 loca-nescent Dosimeters (TLDs) 89 tions; Pressurized ion ,
locations and 19 quality chambers increased to 50 control at present; Pressurized locations ion chambers - real time moni-toring: 16 locations at present AQUATIC MONITORING Present TMI-1 Program Recommended Program
- 1. Wate' Column: 9 locations at Increased to all effluent .
prese. : I-131 bi-weekly, gross pipes, all public water supply beta, gamma, H-3, P-32, Fr-55, intahes, and intakes for crop gross alpha, all monthly; St-89, irrigation. Continuous sampling St-90 quarterly checked every two weeks for iodine and monthly for other radionuclides, in addition to that at left.
- This table is based on the TMI-REMP, a recommended program in Patrick and Palms, 1987, and additions by the TMI Public Health Fund Scientific Advisory Board. .-)
- 66 -
Present TMI-1 Program Recommended Program
- 2. Groundwater: 25 indicator wells 25 indicator wells as per the as per the clean-up monitoring clean-up monitoring for gross for gross alpha, beta; isotopic alpha,-beta; isotopic gamma '.
gamma emitters; tritium and emitters; tritium and Sr-89 Sr-89 and Sr-90 and Sr-90
- 3. Sediment from Shorelines: 2 Increar,e to 5 locations: twice locations at presents semi- annually at early summer and '
annual gamma scan, St-89, low flow for samples at.left. Sr-90 In addition monthly samples of i floc on each side of the river, - near any water intake, and downstream of effluent outfall. Sample for I-131, gross alpha, , beta, and all indicated iso- , topic gamma emitters.
- 4. Aquatic Vegetation: 2 Increase to monthly periphyton locations at present sampled samples from just above plant, ,
bi-annually. control area, ends of all ef-fluent pipes, and all water intakes downstream unless a no effect area is established. J Sample for I-131 and isotopic , gamma emitters 7 5. Aquatic Insects: Not sampled Sample insects at a control r l presently area just above and below the l plant and at Columbia-Wrightsville area. Sample for
- isotopic gamma emitters twice a y>ar
- 6. Fish: presently take one sam- Increase to: 3 collections 1
pie per season of preditors annually at early spring, low and bottom feeders flow / warm weather period and-autumns of herbiwores, detriti-vores, omnivores, and carni-vores. Collect in a control , area determined to be where fish will not migrate as well j as above and below the plant. ; l Monitor for isotopic gam.ma 1 emitters and St-90 and 89 ; i i
! e i i 1
1 l
TERRESTRIAL MONITORING i Present_TMI-l Program Recommended Program
- 1. Food Products: ;
(a) Fruits and Vegetables (a) increase to: fruits and
- presently one sample at harvest vegetables and leafy vege-time of fruit, vegetables,-and tables from four farms and.
broad leaf vegetation. farm stands, and an annual - survey of farm gardens,in each , of sixteep sectors out(to five I miles as well as the fertili2- - ~ er, sprays, seeds, and seed- ; lings of gardens at the plant fence in the highest areas of ' concentration. Harvest air from these gardens and sample corn, tomatoes, beans, pota-toes, head' cabbage, carrots, l " lettuce, Swiss chard, and cabbage leaves. Sample roots, stems, leaves, and fruits ; (b) Domestic Animals and Their Increase to cows within S-7 l Products: (1) milk currently miles at the highest air cont ' sampled at 8 sites on a semi- centration areas, in all 16 monthly basis sectors if possible, and near r to plant farms wil5 single cews; in addition, goats in the ; same pattern as well as dair- ; les. Additional samples of ! pasture grass for goats and ; (2) eggs -- not currently cows; (2) eggs -- from none ' 1 sampled to:- five farms in highest air concentration areas and con- . trols in least prevalent wind (3) beef, sheep, swine - direction; (3) beef, sheep,, not currently sanpled swine--from none to: slaughter houses that take animals from ; the local areas l l
\
.q I
u { ] Present TMI-1 Program Recommendef Program ) Non-Domestic Animals Eaten
~
(c) (c) Program to consist of- ! Locally: none presently sampled (1) rabbits, squirrels, and woodchucks at 2 locations in
- highest . air ~ _ concentration i areas, 1 control in a crosswind-E areas (2) - deer 2 . samplea f rom l local wardens;' . (3) birds,
- grouse, quail, pheasant, ducks, geese -- one;or more locations
- of nesting sites o r., known feeding areas in the ! plant's
~
vicinity and control area, 14-15 miles from the plant in a crosswind location, j 2. Estimates of Accumulation: ! (a) natural bioaccumulator (a) Program to consist of plants: none presently sampled lichens (evergreen-moss) at plant boundary, 5, 10, and 50 j miles, and predicted locations i of hot spots and plume touch- ! downs, and control areas with j - little chance of contamination; 1 . ! (b) soils sampled twice per year (b) increase to a grid de- { } veloped for nearby forests, pastures, and gardens. In addition, analyze humus, litter, and sod from nearby forests twice per year for gamma emitters. k
*N Current Practice 3 The USNRC Regulatory Guides 4.1 (USNRC, 1975a) and 4.8 i (USNRC, 1975b) set forth the regulations for monitoring radionuclides in the environment. Under these regulations ;
radiological monitoring programs must provide data on measu6able levels of radiation and radioactive materials in the s)te ; environs. These measurements should be used to verify the i effectiveness of in-plant effluent treatment systems. In ; addition, as a result of the TMI accident and the on-golig clean i up program, supplemental routine monitoring is done at TMI-1 by the NRC, the EPA, DOE, and the Commonwealth of Pennsylvania.
.- t The NRC thermoluminescent dosimeters (TLD) network .,
consists of two concentric rings of TLD stations arour.d the nuclear plants one ring between the plant site boundary and 1.0 t t to 2.0 miles out; a second ring from the outer boundary of the first ring to 5.0 miles out. Several ether TLD stations are i j located beyond the five mile ::ene . The URC network usually j contains 16 stations per ring (one per 22.50 degree sector) and . requires overlap with five utility TLD stations and the nearest i residence. The network provides up to five TLDs per population 4 I l i 2
- 3. (This discussict. is based on Section E.1 Present Direct Radi-
- ation Monitoring by Agencies in Appendix E, "Atmospheric Monitoring" by William Harding in Light Water Reactors. Moni-l toring and Management In-Plant and Environmental For Three i Mile Island, Academy of Natural Sciences, Philadelphia, s i
2 1987.) '
-d 1
i e . l 1 i d center and three stations positioned approximately fiftecn to- , l twenty miles from the plant. - ) Direct radiation monitoring devices under EPA authority , pre:dently include TLDs and a Reuter-Stokes RSS-10ll SENTRI system j 4 of lowlevel (1 micro-R/hr to 100 milli-R/hr) pressurized ion chambers (PICS). Due to EPA budget cuts, the, remote on-bine ! I f j portion of this system was discontinued in July 1987. l The Commonwealth of Pennsylvania administers its nuclear f facility environmental monitoring responsibility through the 1 r j Bureau of Radiation Protecti0n (BRP) of the Department of l Environmental Resources (DER). The bulk of the routine 1 i j monitoring is presently based on' a TLD and air sampler network '
. 4 which includes 18 TLDs (to cover the 16 sectors and the ,
i population centers) and three air sampler locations. i i l The present monitoring scheme provides for off-site l 1 , i gamma-spectroscopy on soil and vegetation at seven sites by the !
- United States Department of Energy (DOE). -
l 1 l ;
- l The State of Maryland does not specifically monitor TMI l f but does monitor certain aquatic biota and sediment associated in 4
f the lower Susquehanna and upper, Chesapeake Day environs. This j t provides another useful TMI related data source. s
..(
1
- ,i
\
i . In total, the scope of environmental monitoring around TMI-l currently done by agencies other than TMI management is quite extensive. It totals over one hundred (100) TLD stations and covers all sixteen sectors and various segments for each sector. The bulk of the monitoring is done, however, using the passive thermoluminescent dosimetry (TLD) devices, most of which are read-out only every three months. , j' Practical Problems With Current Practice Gaps in t!.9 environmental media sampled and problems with the performance of direct radiation monitoring networks are two continuing problems associated with TMI-1 and similar off-site monitoring programs. 4,
- Potential Missino Pathways Current U.S. regulations do not require monitoring of l
Carbon-14 releases. Current Swedish and German regulations I l (Till, 1987) recognire the long-term impact of C-14 on the biosphere and thus require limitations on the amount released and monitoring to verify compliance. Cel4 emissions arer one missing j pathway and C-14 monitoring should be included in U.S. nuclear plant monitoring systems. Comparison of the current TMI sampling program with the San Onofre and Susquehanna programs reveals many other unmonitored pathways. These include local food sources; -
non-domestic game animcis; forest soils and humus; lichens and mosses; floc and periphyton. Currently, at Susquehanna, the utility maintains a series of off-site test vegetable gardens used in a regular sampling program. The sampling of such media should be employed at TMI and expanded as it provides important back-up data for dose assessement, regulatory compliance, and potential ar'idents. . Direct Radiation Monitoring Networks Studies of direct radiation monitoring networks similar to the TMI-l program (Maeck, 1982 and Franke, 1987) examined the ability of a fixed real-time monitoring system to detect and
~
quantify releases; assessed the uncertainties asso :iated with ' . estimating unmonitored releases; and reviewed the availablility, cost, and instrumentation requirements for a system. The sensitivity of a fixed atmospheric monitoring system is dependent upon the ability of the fixed devices to measure readioactivity in their immediate vicinity, and the number and - location of the devices. The effective monitoring of radioactivity by fixed devices is also dependent upon their threshhold of detection and whether monitoring capability is continuous or intermittent. Because ground level concentration of radioactirity from a plume at any given distance from the emission point may vary widely, depending on factors such as wind
-b l
1
- l
/ . .
speed, atmospheric stabill'y, t and terrain, the appropriate number and effective placement of monitors becomes problematic. . Uncertainty of Detection Using uncertainty estimates, based on the use of simple error analyses of the meteorological expressions required'(to describe plume shapes and atmospheric transport, a 1982 study by Maeck concluded:
- 1. In a ring of detectors around a nuclear power station the number of stations required for two detectors to provide information within a factor of 5 of each other can be up to 50 or more for one installation. .,
- 2. The use of short-time (every 15 min.) data from a fixed off-site monitoring system to project downwind dose rates
!s a complex process carrying an uncertainty factor of 10 or more. Complex terrain further complicates such projections.
- 3. The use of a fixed off-site monitoring system to estimate the amount of release of unmonitored radionuclide from data on the amount of a release of monitored radionuclide is highly quertionable and carries an uncertainty factor of between 25 and 50.
31
- t>
l
- 74 -
9
Figure 3: Locations of air iodine and air particulate monitoring, around TMl, before the accident in 1979. (the following locations.are outside the map area: three Met-Ed and one DER) 5 - g 0 : e . O [ 1-F ' 2- j -' 5 i i , , i i 5 3 1 1 3 5. KM o MET-ED aDER l From Franke,1987 I I s
-b i
Figure 4: Present (1985) locations of air iodine and air particulate monitoring, around TMI: (the following locations are outside the map area: one DER, three GPU, and one EPA) s 4-3 . 8 .
'- ' +
3-
+ ~
2- s 1-
)
E o n (- 1
- 'i 4 - .
O : 1- . 2- L 2 g 5 i i i i i i i i 5 3 1 1 3 5 KM OGPU + EPA O EPA hi voi aDER x NP.C From Franke,1987 t
%1
C e 1 l l 1 Figure 5: Locations for monitoring of air iodine and air particulates in a hypothetical improved system. 10 o u 9-0 0 6 0 3-2-o o . g a c 1- O E o g - C - M - o c9W 1- / o . 2-0 0 " O C
; D 3- -i
. C - O 4_ 5-a U 0 Yh 6- o s%t. 7-8- o a
- g. .
n m y ,. 10 , , , , , i . . , , , , i , , in i 10 8 6 4 2 0 2 4 6 8 10 KM Note: For this illustrative analysis, the assumption was made that monitoring stations are evenly distributed on each circle. As this assumption leads to some stations being located in the Susquehanna river, some adjustments concerning the actuallocations of the systems must be made. The results of the sensitivity analysis would only be slightly cl*Jnged. , From Franke,1987 s h
To illustrcte these uncertainties Franke (1987) inves-tigated the capability of TMI past, present, and potentia'l atmospheric monitoring systems to detect releases. For hypothetical one hour releases of I-131 and gamma radiation, he divided meteorological conditions into six atmospheric stability classes, six wind-speed categories, and sixteen wind direction sectors. He classified TMI weather data for 19.S3 into the h76 , (6x6x16) available categories and he postulated a 50 meter release height and a straight line Guassian plume model with West German Regulatory dispersion coefficents. I-131 Monitorinc Problem: The first monitoring test was for I-131 in air. For the current TMI-l monitoring system, the I-131 threshholds of detection (pCi/m3) were: GPU C.34; EPA 0.90 * . for 833 min; EPA hi vol 0.55; DER 12.0; NRC 12.0. The air
' concentration per unit of release was calculated for each of the 576 weather categories and a maximum undetected release ,
calculated for each category. The hypothetical improved TMI-1 system devised by Franke for I-131 monitoring around TMI-l featured 64 monitoring stations in four circles with radii at 500 - meters, 2000 meters, 5000 meters, and 10,000 meters. Figures 3 to 5. Franke plotted a cumulative distribution curve for each monitoring system showing the frequency in hours per year, wi.th which various levels of release could escape detection. The results indicated that there was no period during the year when +
- 75 -
o . Figure 6: Undetected 1-131 releases: ' Frequency of I-131 releases, with a duration of one hour, which could go undetected by air monitoring 9,000 E 8,000-y 7,000 -
>- e Oh N 6,000 -
5 d ea! cz: i
@$ 5,000-GB l 58 l 4,000-
@ g
- 3 g
- u. W O$ 3,500 $ ,
i g 3,000- l , s5 l 8a i
@U 2,000 -
1,700 1 1,000 - l l l 0 , , , , , , 6 4 la aa kddd$S .s 2i i23 0.000001 0.0001 0.01 1 100 10000 1000000 - RELEASE (in curies); Ouration: I hour O BEFORE 1979 Q PRESENT SYSTEM 6 IMPROVED SYSTEM i How to read this graph: The horizontal scale indicates one hour releases of I-131 which could go undetected, The ! associated value on the vertical scale indicates the number uf hours per year during which it I is possible that an undetected release of this magnitude or even greater could occur. The smallest undetected release (most favorable weather condition) can be determined from the point where the graph departs from tile horizontal after leaving tiie vertical axis (top left), The largest undetected release (most unfavorable weather condition) can be determined from the point where the graph intersects the horizontal axis (bottom right), From Franke,1987 s I
Figure 7: Locations of thermoluminescent dosimeters (TLDs), around TMI, before the accident in 1979. (five locations are outside the map area) c - 6 s' .: V S' ht s \ s
\ '
5 S $ b KM 950* es @ #' # '
% \
t g I f Faure (TLDs)around 8: Present TMI(1985) (20 km quadrant) locations of thermolumines (the following locations are outside the map area: 12 NRC, 8 DER, 26 GPU, and 13 EPA) 10 - 9 - .
? ; 8 -
U 7 ^ 0 o C
. O o N o
6 - ' S g 5 - o O o o 4- 3 gg a
^ E D 3- '
0 60 ' 2- 3 g 0 ( a 1-0: 0 llE 0-k* E 1- O o O 2- _ g o I 3- .oW ^ t . L o a 4- , 5- o k a 6- V o 7-
+ .!
t ' o 101 ,,, a Ng 10 8 6
, , , , , , , , ' ',*[ , y' ",*T'-
4 2 0 2 4 6 8 10 KM aNRC + DER oGPU aEPA From Franke,1987 s,
- Ol
l l Figure 9: Present (1985) locations of real-time (SENTRI) gamma i monitoring systems around TMI (two EPA stations are located outside the map area) i ! 5- 9 . 4- u o
' U 3- ,
a \ 2- " 9 1- (', ' b 2 0 ' 01 -
. t - -
' i u , z.,
)
C ' a o 2- , O N l 3- a o U 4- - 1 O s 5 , , , , 0 5 3 1 1 3 5 KM aEPA 0GPU From Franke,1987 l l
Figure 10: Locations of TLDs in a hypothetical improved system. 10 u C u D U 9-O U 8-' N 7- o . O h 6- . g 5- N 4- g -
. o 3- Ng oog C
2-J 1- O hh1 C. ' O 0 E lE 0~ o O # goU *3 g O - a - 1- 0 iy O 2-] % a nO J' C - , 3-4- D - D 5-o o o 10 , i , , , , , ? i :: , 7ii,,i7 , 10 8 6 4 2 0 2 4 6 8 10 KM Note: For this illustrative analysis, the simplified assumption was made that TLDs are evenly distributed on each circle. As this this assumption leads to some TLDs being located in the Susquehanna river, some adjustments concerning the actuallocation of the TLDs must be made. The results of the sensitivhy analysis would only be slightly changed. From Franke,1987 s
..h
l l l Figure 11: External gamma doses: . Frequency of the ratio "maximum potential gamma dose rate / maximum measured gamma dose rate" for releases with a duration of one hout The real-time gamma monitoring stations are compared with different sets of TLDs. The following systems were compared: 1) the present (1985) real time (SENTRI) gamma ionization chambers; 2) 20 TLD stations in effect before and in the first days of the TMl accident: 3) the present (1985) array of 193 TLD stations: 4) an improved system of 160 TLD stations which are systematically distributed. 9 11 6,000- i b I 8% a{ E !! 7,000-aa i ' gj 6,000- 1 a3 23 ' 5,000-h 45 se - .
~Q o 4,000 -
ED g g
$$ 3,000-DE Zc
$y Op 2,000 -
E4 w0 1,000 - L L 0- - i - y - y u'. , - .p - o" 1 3.16 10 31.4 100 316 1,000 RATIO (of maximum / treasured dose)
. OSENTRI + 20 TLD (<1979 Q 193 TM%Ifay) 6160 TLD How to read this graph:
The horizontal scale Indicates the ratio "maximum potential gamma dose rate / maximum measured gamma dose rate" during one hour releases of noble gases. The associated value on the vertical scale indicates the number of hours per year during which it is possible that an undetected release of this magn'tude r or even greater could occut The smallest ratio (most favorable weather condition) can be determined from the point where the graph intersects the vertical axis (top left). The largest ratio (most unfavorable weather condition) can be determined from the point where the graph Intersects the hortzontal axis (bottom righ'), , From Franke,1987 e
the hypothetical improved system would fail to detect a one-hour release of 0.01 Ci of I-131. However, the current system woul'd fail detection for approximately 1700 hours / year; and the pre-1979 TMI-l system would fall detection for approximately 3500 hours / year. The detection capability of the improved potential TMI-1 system is due to the increased number of monitors, improved distribution, and lowered detection threshholds, Figure 6. f . Gamma Radiation Monitoring Problem: To set up a test model for detecting gamma radiation released from noble gas, Franke examined the sensitivity of the four monitoring systems at TMI -1; the pre-1979 TLDs; and the present TLDs and SENTRI pressurized ionization chambers. For the hypothetical improved system he used 160 TLDs in five circles with radii of 100 meters, * . 300 meters, 1000 meters, 3000 meters, and 10,000 meters, Figures t 7 to 10. For this model, the maximum actual gi'ma dose rate in the area surrounding TMI was compared with the maximum gamma dose measurement permitted by a given measuring system. It was assumed that the dose-rate exceeded the threshhold of detection for all weather conditions. Calculations were directed at assessing the effect of the number and location of monitoring devices on the sensitivity of a system. Calculations were done for each of the 576 weather conditions and air concentrations at each monitoring location calculated. The ratio of "maximun 1 actual air concentration to maximum air concentration at each ~ device site" was determined and assumed to be equal to the ratio of "the maximum potential gamma dose rate to maximum measured
. _'t l
l
c.
" for all 576 weather categories. A cumulative gamma dose rate distribution curve was plotted for each monitoring system, Figute - 11.
Results showed that the SENTRI system sensitivity is similar to the pre-1979 TLD array and for the one hour release period the maximum potential gamma dose rate could exceed $ hat measured by the SENTRI system by a factor of up to 170. The equivalent factor for the TLD system could not exceed 18, and for the improved potential TLD array it could not exceed 6. Costs Evaluation of the cost and performance characteristics ., of available pressurized ion chamber instrumentation shows that: (a) instrumentation costs are relatively fixed, but the installation costs are variable; (b) the capital cost per monitoring station ranges f rom S25,000 to $65,000; (c) depending upon site characteristics, the capital cost for a 32-station system may exceed $1,000,000 but only provide data with. uncertainty factors ranging from 10 to 50; and (d) the placement of a simple 8 to 16 station (5500,000) detector system in proximity (0.5 mi) to a r eac.to r may not provide reliable information in an emergency because gaps in detector spacing could permit the plume to go undetected. In general, the effective emergency monitoring capability of a fixed station (16-32 units) emergency monitoring system is highly questionable, s.
,_d
l Other important considerations affecting emergency real-time on-line transmission of data are weather conditions and l phone service. At TMI both the EPA and GPU Reuter-Stokes SENTRI systems of PICS linked by phone lines to remote locations have experience problems. The availability of EPA remote data from any one monitoring location is approximately 75 percent; there is only a 25 percent chance of having all 13 EPA units on line a'nd , working simultaneously. This is due to telephone line malfunctions and failures of circuit board components. Circuit board problems are caused by rapidly changing weather, lightning, and extremes in temperature and humidity. They are aggravated when the four telephone companies that supply the lines for telemetry have circuit problems (Kirk, 1987). Weather resistant redundar.t systems and satellite telemetry are alternatives, which . cost substantially more than the SENTRI system. Procosed Imorovements in Current Environmental Monitoring State-of-the-art methods of monitoring and specific data on the TMI site and situation are the bases for proposed improvements in the current TMI-l environmental monitoring program. Under. routine operating conditions, environmental sampling should overlap with organisms, sub groups, individuals, and environments analyzed in the end-effect studies recommended in Chapter 5. During an accident the dispersion of the radiation fields will determine priority sampling locations. s.
..h
TMI Environmental Considerations For Monitoring Atmosphere At TMI, the Susquehanna River flows across broad valleys and high parallel ridges. Such ridges cause the development of a different pattern of air movement than would occur in the coas,'tal or Midwest plain areas. Wind speed is important in determining if air releases from the TMI plant will hover over the ridge, or move up and down the river valley. The longer a release moves in a given direction without dispersion, the greater the deposition of particulates and the effects of fallout on vegetation and organisms, particularly man. (Patrick, 1987) Inversions are a definite problem in many areas and are particularly severa in the Susquehanna River Valley at certain times of the year. They prevent the air from rising, thereby concentrating any releases from the plant. The dose to man under such conditions is greater than when there is free upward movement of the air. Use of the AIRDOS-EPA and GAS?AR air dispersion models for the TMI site have provided insights for the location of monitoring points, and critical radionuclides to monicor. Beyond the 7.5 mile radius, the maximum air concen-tration (X/Q) is reduced by a factor of approximately 8; at a .l
. \
distance of 10 miles, there is an order of magnitude decrease in s
the calculated X/Q's. These factors suggest that it would be cost-effective to keep routine monitoring stations within a ten mile radius of the plant; a_five to seven mile distance would-assure the optimum use of funds available for sample collection and analysis. (Till, 1987) However, odd effects can still occur in the valley's even greater than 10 miles away. Thus, there should be some instruments placed at greater distances. f . Maximum air concentration for radionuclides occurs 0.5 miles SSE of the TMI plant at a point located over the Susquehanna River. The closest inhabited points where X/Q's are greatest occur near the shore of the river between the NE and SE compass directions. This (0.5 miles to 1.5 miles) is the most likely area in which to establish monitoring sites. For dis- ' ,- tances beyond a 1.5 mile radius of the plant, the maximum X/Q values shift to the WNW and W. The terrain covered by these sectors is rolling agricultural land with valleys and ridges. Therefore, this area should oe strongly considered for placement of routine monitoring stations for measuring air concentrations of gases, determing deposition of particulates, and analyzing - food products. Beyond a radius of 7.5 miles, maximum air aw concentr'ations occur in both the WNW Y6'3SE directions (Till, 1987). Although maximum air concentrations to the north of the plant are among-the lowest, it is essential to incorporate this area of highest population density as pa'rt of the monitoring 4
. o program, checking primarily for air concentration of radionu-clides and deposition rates (Till, 1987).
The AIRDOS-EPA computer code does not allow for the possible effects of terrain on the calculated X/Q. Although the roughness of terrain may affect the calculations of X/Q to some degree, it is not likely that these values would be affeped significantly within the 7.5 mile range. Beyond the 7.5 mile radius, the roughness of terrain becomes more pronounced including mountain formations to the north, southwest, and west. Within the 7.5 mile area, the most significant clevations do not 'ie within the areas best suited for monitoring. The possible exception to this is Hill Island, which lies to the north between the TMI site and Middletown. Although this terrain , could create a difference between the observed and calculated ! X/Q, the effect would be minimal (Till, 1987). l l Based on the maximum potential total dose, routine - l emissions are low and measurements of specific plant generated radionuclides in the environment resulting from normal operations . will be difficult (Till, 1987). Radionuclide pathway analysis indicates that direct plume overhead dxposure is dominant, - contributing more than 97 percent of dose to human organs. This l finding shows that monitoring environmental nadia for surface exposure is important (Till, 1987).
-y
. e The potential for exposure to the public is greatest from the noble gases Kr-85 and Xe-133. , This is significa'nt because measurement of noble gases released during routine plant cperations is expensive. However, measurements of noble gases at the source (in effluent) can be done on a continuous basis with an effective effluent monitoring system. Therefore, it is recommended that these radionuclides be emphasized in i'the -
proposed in plant effluent monitoring system. (Till, 1987) Rivers and Ground Water To monitor a river it is important to determine the aquatic pathways leading to man. 7. n the Susquehanna River, the main pathway is fish. It is also important to determine the dose ' . to humans from drinking water or from exposure during water recreation (Patrick, 1987). To do this correctly the flow and characteristics of the current pattern of the river must. be understood. Flow patterns determine the length of time that a given intake of water might receive a contaminated slug and the rates of sediment deposition (Patrick, 1987). - The Susquehanna River in the vicinity of the TMI plant is an old, broad, shallow river with consolidated bed material. In the past the river had swifter flow and carried a high sediment load. This has resulted in a braided pattern ~' interspersed by small islands throughout the river channel. This s
-t
. o e
flow pattern is very different than it is in a free-flowing river with unconsolidated bed material (Patrick, 1987.). The braided flow pattern creates many small channels. Some of these channels move swiftly, others form backwaters and small pools. It is relatively easy to know how soon a contami-nated slug of water will reach an intake in a. swif tly flo$ing , channel, but the complicated pattern of backwaters and shallow pools in braided streams makes it much more difficult to deter-mine when the last of water bearing radioactivity has passed the intake. Flow analyses must be done at both high and low flow rates (Patrick 1987) Sediments and Soil , It is well known that sediments sorb large amounts of radionuclides. The fine particles, particularly colloids, in sediment are most active in sorption. Colloids iray be organic or j inorganic in composition. They typically move with the water column and only settle out where the water is extremely slow or . scarcely moving. They may be aggregated by changes in the ! I chemical composition of the water or they may sorb onto larger ) particles (Patrick, 1987). Common deposition points for sediments are behind dams, and behind eel weirs. Sediments also settle out in small pools and backwaters. If the oxygen levels or the redox potential are 4
-t o e greatly lowered .in backwaters and pools, sediments may release substances that were sorbed onto their surfaces into the wate'r column, thus increhsing water column concentrations of radionu-clides (Paurick, 1987). To determine toxins or radionuclides in water, it is most important to analyze sediments because of the transfer that often exists between sediments and the water column (Patrick, 1987). [
Soils are the depository of atmospheric fallout; they contain many naturally occurring radionuclides. Types and amounts of radionuclides in soil are dependent upon the soil composition. The extensive use of the in situ method for determining radioactivity in leaf litter, humus, and tr.e top few centimeters of soil is an excellent procedure for quickly + , determining radioactivity over a large area. Following an accident, soil should be useful in determining the extent of the area af fected by fallout. Portable Gri. cou*ter systems can be helpful in locating areas and objects f rc:s ;hich to collect soil samples (Patrick, 1987). Soil coring permits the study of soil characteristics for specified distances beneath the earth surface. Because soils vary in structure and particle size, it is important to determine the soil profile and the root structure of associated plants to determine a suitable coring depth. The dust should be sampled to determine airborne and resuspended radioactivity. Core. samples should be taken at various depths to determine any buildup of *
._t
- 84 -
long-lived radionuclides, the potential for plant uptake, and possible future resuspension as dust (Patrick, 1987). Organisms Organisms selected for monitoring should bioaccumulate large amounts of radioactivity and be native to the plant a; tea ' and a component in the human food chain. Selected organisms should also: grow in all seasons; have large surface-to-volume ratios; and have surfaces exposed to the air the water current (Patrick, 1987). Recommended terrestrial bioaccumulators to be monitored in the Three Mile Island region include folious lichens; recom- , mended aquatic bicaccumulators selected for monitoring are periphyton (Patrick, 1987). Periphyton is an alga composed mostly of diatoms. It is a principal base of the food web in most rivers and lakes. Periphyton are collected using artificial substrates, which are , very effective for accumulating and growing a periphyton community typical of the area of environments whe're traps are placed (Patrick, 1987). In a temperate zone river, such as the Susquehanna River, two weeks is usually sufficient for periphyton to grow to create a large enough sample to provide multiple slides for analyses (Patrick, 1987). s,
..D v
Because periphyton are covered by a carbohydrate col-loidal material, they quickly sorb large amounts of radionuclides onto their surfaces and are excellent fc- immediately indicating the presence of radioactivity. They are also excellent for determining the presence of radionuclides that could cause poten-tial long-term risk to humans, especially via water ingestion. Continuous cell division provides new surfaces for sorptkon. This capability for sorption permits these diatoms to accumulate many thousand times the ambient radionuclide concentration. The exact amount of concentr'ation depends upon the season of the year, the type of radionuclide, and the amount of nonradioactive chemicals present. Periphyton were able to pick up, better than any electronic instruments, the radioactive fall out. from Chernobyl in the Suaquehanna River. (Patrick and Palms, 1987) * , Periphyton should be collected monthly in routine analyses (Patrick, 1987). Variations in samplers and sampling technique could result in up to, but not more than, a 33% variation in findings. . Analyses should be conducted for beta and gamma emitting radionuclides and should be modified to include - alpha emitters (e.g., Pu-239 (24,390y) Am-241 (475 yrs), and (2.14 x 10 6y)). to #Yl$h
~
Np-237 The dose or other aquatic organisms and, indirectly, the dose to man can be predicted (Patrick, 1987). Entire groups of organisms, foods, and food sources must be sampled to assure that dietary pathways have not been *l
.d 1
. s contaminated by radionuclides. The uncertainty inherent in the mathematical pathway models suggests that the closer monitorin'g is to the exposed population's, food, air, and water, the more accurate the dose projection.
The recommended monitoring of radionuclide accumulations in soils, sediments, and organisms highlights the lack'Iof attention paid to environmental accumulations of radioactive materials by t'ae U.S. Nuclear Regulatory Commission. Estimates by England (l!187) show that the routine operation of nuclear power plants in New England has resulted in environmental accumulation of several types of radioactive materials. Routine use of the recommended accumulation monitoring program at all U.S. reactors could sharpen such estimates. However, these , estimates suggest that the U.S. Nuclear Regulatory Commission should seek to enforce ceilings on the environmental accumula-tions of radioactive materials with long half-ltves, not their annual discharg es . If 'adioactive emissions persist for years, decades or even centuries within. the environment, then even modest reducticns in annual discharges may not be sufficient to . prevent an envi.ronmental build-up of those materials over time. Only if annual discharges of such material"are reduced sharply - from year to year can we prevent their accumulation over time. (England, 1987) s
.g
- 97 -
o . 1 Improvements In The Fixed Off-Site Monitoring Network 4 Underlying all recommendations is the recognition that an effective, comprehensive off-site monitoring program for atmospheric radioactivity requires: (a) a careful mix of instru-ments and technologies; (b) redundancy in programs, instrumenta-tion, and systems to assure comprehensive checks-; (c) knowled'ge- . able site-specific instrument selection; (d) careful calibration and maintenance of all instrumentation, and (e) a competent management and staff who are dedicated to serving the public. Transluminescent Dosimeters (TLDs): TLD stations should carry multiple (at least four) TLD's * . to record for long and short accumulation periods in the event of v. an incident. The Fund's advisors concur with Harding, while the Academy recommended only two per station. Installation of more TLDs per station, the number of
- stations increasing by radial distance from the site, will yield -
an increase in coverage and precision and is recommended. It is proposed that a variety of measuri q instruments be deployed in i sufficient numbers, and in such placement, to assure coverage of ' the ambient radiation field.
- 4. (This discussion is based on Sections E.9 to E.13 of Appendix s E, "Atmospheric Monitoring" by William Harding in Lighq, Water Reactors, Monitoring and Management In-Plant and Environ-mental Study For Three Mile Island, Academy of Natural Sciences Philadelphia, 1987.) ' ~}
o . The NRC recommends that a total of 40 TLDs should be
.used. One TLD should be located in each of the 16 sectors 'at approximately the site boundary, one in each of the 16 sectors at a radius of four to five miles, and eight more at locations "of special interest" (NRC, 1979). In addition, the NRC proposes placement of an. additional 16 independent NRC TLDs in a ring at the three mile radius, one in each sector. This would proyide adequate coverage only under unstable atmospheric conditions (Classes A-C, perhaps); coverage using four TLDS per sector seems to be required for coverage in stable conditions. This would quadruple the number of TLDs currently deployed.
For accident conditions, supplementary monitoring should be provided using portable instruments. Althcugh these instru- , ments measure dose rate instead of dose and usually provide only qualitative information, they often provide the best timely information on dose rate at a given site. Each TLD station should have several direct reading dosimeters (DRDs) to determine if instruments should be replaced . more frequently during an incident. At each station some TLDs should be configured to provide a qualitative indication of beta dose. N
*(
Pressurized Ion Chambers (PICS) At TMI-1, an on-line real-time PIC network instruments should be set up and linked with the radiological dispersion modelling capability, as has been done at other nuclear power facilities. The 'use of the PIC network in this feedback or predictor-corrector mcde can provide a more accurate real-t$me - estimate of environmental radiation impact. The PIC network must have at least six portable units with radio-frequency capability and a backup communications link to the central computer facili-ties. Associated PIC software should be developed to show a plot of the standard deviations of PIC measurements over time which will indicate radioactive plumes in accident situations. In-addition, all PIC stations chould have at least two types of ' TLDs to provide independent and back-up radiation data confirma-tion. Spectrometers Several (at least three) prrtable gamma spectrometers ' units should be available for emergency use. These units should be capable of working in high radiation environments. Air Samplers Several (at least three) portable air sampling units should be available for incident use. Fixed air sampling units' ' 90 -
I should be provided downwind of the plant and at least three fixed and several portable air sampler units should be provided f6r surrounding populatn centers. Procedural Changes Inhisreviewofinstrumentation_fortheAcademy(19k7), , Harding made the following recommendations: (The Fund concurs wi.th Harding. The Academy's main report only calls for back-up systems not immediate readiness.) (1) Determine if all monitoring instrumentation will function properly from cold start-up.
~.
9 (2) Train all monitoring personnel in well-established hardware, software, and firmware audit procedures (including test I data and test signal generators) and apply these procedures (periodically, at random, and immediately after the intensive i phase of any radiation-related incident) to all radiation monitoring equipnent, associated readout, and data reduction - devices. '1 (3) Develop a prede;v: mined survey network staffed by 4 voJunteers (police, firemen, and emergency personnel) supplied with ionization chamber survey meters and two-way radio communi-cations. .
e (4) Establish a portable microcomputer system for I emergency survey teams with the ability to send real-time data to a central computer system. (5) Develop a real-time capability for air sampler and PIC data. Feed the data back into the radiation dispersion program for immediate radiation dose impact estimates. f (6) Place the on-line real-time monitoring system with dispe: . on model feedback link in a basic grid 2.5 miles f rom the nuclear plant site center. It should respond to class E meteoro-logical stability conditions, and 50 percent plume centerline concentration. In addition, place two or three monitors at or near area population centers; and place two portable (radio- ' . frequency linked) monitors in alignment with with monthly and seasonal preferred wind directions. (7) Develop a flexible mobile emergency monito ic.g s.ystem which may be cart.ed in several types of vehicles (boats, 1 vans, light airplanes). Install monitoring instruments in - aircraft so that they can be used to map out the fallout pattern in case of'a major accident. Sucri aircrNTY>ere u' sed ef f ectively during the 1957 Windscale accident in England.
.t
+ ,
CHAPTER 4: REDUNDANCY AND BACK-UP REQUIREMENTS: QUALITY CONTROL e l 4 l 4.
.)
d 1
o . Back-Up and QualJcy Control Concerns A compreheneive radiation moni'toring system must: ; provide adequate coverage for assessing actual or potential doses to a population or sub populations; demonstrate compliance with regulations; and provide warnings of accident conditions. The~ recommended in-house and environmental pragrams must incl,ude
.. j back-up capabilities to achieve these goals even in the event of partial or complete shut-down of the plant on loss of electric power.
The recommended quality assurance program requires a i wide variety of sampling to mitigate for unforeseen events and uncertainties inherent in dose calculations and environmental e
'l dispersion models. The suggested program additions and improvements overlap and complement each other, or fill existing measurement gaps. The history of reactor operations suggests a t need for this redundancy to accomplish quality assurance objectives.
The recommended overlap of measurements is a safety i procedure to guard against unmonitored pathways or partial degradation of the monitoring system. However, the overlap also produces interdependence among the various types of monitoring. Degradation of a portion of the system can affect other . l measurements. Therefore, chutting down the reactor should be required in the case of system malfunctions that hinder adequate , monitoring. -$ i l
, _ _ ..~
f - 1 Assessment of Actual or Potential Doses to Groups and Sub-Populations l Both in-plant and environmental measurements should con-tribute to radiation and radionuclide dose assessment. Together these measurements could provide a well integrated radiation t warning and assessment program. In-plant Monitoring : Gaseous and liquid effluent monitoring systems should provide an on-line real-time warning for planned and unplanned i radionuclide releases. Recommended improvements in monitor ; placement, frequency of sampling, and equipment hardening (i.e., equipment resistance to high temperatures and radiation fields) , would upgrade plant capability to detect problems before radio-active releases could reach the environment. Plant personnel or an independent agency should integrate this in-plant data with , the on-line meteorclogic data, site-specific meteorologic inter- ; pretations, ar.d the dose assessment models to obtain an accurate estimate of potential dose risk and affected populations and . - ecosystems. > Fixed Off-Site Network Monitorina If implemented, the recommended increase in TLDs and PICS, and the improved placement and remote data transmission capability, would provide adequate dosimetric assessment. The s
-t I
recommended TLD network has retrospective value, while the ion chambers (PICS) would provide real-time on-line warning and dose assessment. Thus, the fixed in-plant and fixed off-site networks . would provide a predictive, retrospective, and real-time popula-tion dose assessment capability. , 4 Off-Site Environmental Monitoring . If a gap in the in plant and off-site fixed systems should permit the emission of undetected radionuclides, the ex-tensive environmental sampling network would provide the neces-sary data on specific nuclides, fallout and emission locations, , and varied exposure pathways. Environmental samplea could reveal which isotope is involved in an exposure. Environmental monitor- ' . ing of bicaccumulators and local food and water sources would provide the back-up data for regarding immediate and long-term presence of particulates, iodine, and noble gas radionuclides. Selected animals and individuals should be monitored for dose assessment. If the frequency and location of blocccumulator sampling were sufficient to provide dosimetric coverage, envir- - onmental monitoring would provide the program with an independent retrospective back-up system that w'ould funIII n without electric power. I i _q i I i
- 9s -
. I
. o Regulatory Compliance L
If the recommended in-plant and off-site fixed networks operate properly, they should provide the first-line monitoring necessary for an effective system. Environmental sampling would serve as an effective back-up system in the event of first-line failure and also would provides site-specific . hack-up and ifata verification. In-Plant Contributions Calculated and scheduled effluent releases cre routinely j monitored and recorded on release permits, when computerized as recommended in Chapter 2, this information would become reference , data for future operations. In addition, the recocnended upgraded real-time, on-line, in plant monitoring capacity should record small transient releases caused by variations in reactor operations. Bioaccumulator Contributionc . The recommended bicaccumulator monitcring of periphyton, , mosses, and lichens on a regular schedule weuld record the
. nuclear plant generated characteristic isotopic ratios and could reveal even a transient low level increase of a single radionu-l clide. The sensitivity of these bicaccumulators also would help
- s
-b 97
-. - - _ - ,m
in the identification of radi~ a tion sources other than the local , plant. ; I Provision of Adequate Warning 9 Warnings may be of two kinds. One is preventive and the' other announces serious current trouble. The proposed comprehen- ! sive monitoring system could detect small changes in operation ! which indicate preventive action. It could also indiccte poten-tial or present serious accident conditions in time to warn the population. Detecting Small Chances In Plant Operation i
+ .
The recommended cloaccunulators and the in-plant system wnuld work together to provide preventive warning. Small . increases of one or a few radionuclides might be barely detected by the off-site TLD and PIC network; but the individbal radionuclide would not be identified, so the slight increase could be interpreted as below permiesible limits and thus of no concern. However, regular periodic sampling of bicaccumulators could indicate that a given increase in one nuclide, even if
- below permissible level possibly resulted from an abnormal J l
con'dition in the plant that might then be elucidated by the in- ' 1 plant monitoring system. If a nuclide vere present in greater i _ concentrations than expected, it might indicate a pending ! equipment failure inside the plant. Immediate maintenance could 9 then prevent potentially 7erious consequences. ' _ _ _ _. __ _ _ . _ - .-~
Accident Warningu The recommended fixed, linked, in-plant and off-site networks would have adequate spatial and temporal coverage to detect serious accident releases. The recommended number of instruments has be:m calculated to provide adequate coverage for monitor sensitivity under a full complement.. of atmosphkric conditions. Depending on the time sequence of an accident, environmental monitoring could play an important role in warning. If an eccident were to evolve slowly, the recommended : t envftonmental safoty net could give adequai.e warning to Sring the situation under control. If an accident were to develop qu'ckly, the environmental safety net would be valuable for post-accident , public health protection because the sampled organisms and food products would help establish levels of contamination. Each portion of the program could function separately and still provide the necessary core information. l 5 Quality Assurance and_ Control Methods Concerns of Quality Assurance and Quality Control A Quality Assurance (QA) program is recommended in order , to validate the effectiveness of the comprehensive monitoring I
- 5. (From Appendix G, "Comprehensive Quality Assurance and l Quality Control" by Abraham S. Goldin .in Li ht Water ,l Reactors, Monitoring and Management for Three _ Mi e Island, j Academy of Natural Sciences', Philadelphia, 1967 forms the ~d basis of this discussion.)
i 99 - J
i
-t program. A good quality assurance progra: would assure that the' correct questions were being asked and that the appropriate instruments are properly calibrated and used to optimum advantage. It also would assure redundancy: the implenentation of supplementary programs which serve as additional protective !
monitoring checks, in case of failure or question about the accuracy of one component. The scope of a quality assurance program to insure ef-fective radiation monitoring should cover verification of: (1) I the precision and accuracy of calibration and sampling measure-ments; (2) location, frequency, and nature of measurements; . sampling p.ocedures and information; (3) calculations; data ; handling (4) and a host of other topics. It would also be ; essential that a quality assurance program include extensive documentation. As noted in EPA specifications: "data must be scientifically valid, defensible, 4.nd of known precision and accuracy." ' The failure to ask the correct questions in a quality , assurance program could result in inappropriate samp1'.ng and review or data. This, ultimately, could result in the inaccurate i
' assessment of situations. Quality assurance must also be incorporated in tha design of instrumentation and the development of dosa measurement systems. For example, not asking the right questions about instrument calibration or site-specific design,
- could have serious implications for nuclear monitoring
- 100 - .
. o capabilities. Quality assurance ultimately must function to keep the comprehensive monitoring program acc. urate and to permit fulfillment of stated goals.
Quality Assurance Program l l
)
A quality assurance program (QA) should be in eff;ect during all phases c[ development of a monitoring program.- As appropriate, all techniques and procedures used in a program (from sampling to measurement, data handling, and reporting) l i should be written as Standard Operating Procedures. Program Components l All quality assurance programs should be documented. Each document must include a policy statement, definitica of l l management structure, quality control guidelines, and procedures for auditing an1 assessing the program. i l A separate quality assessment project plan should be , developed for each type of monitoring procedure - (QAMS, 1980b). These project plans should include s description of the project including pucpose, scope, and design. Each project plan should also include a policy statement thatt establishes the framework within which the QA activities are to be accomplished, states ' project goals in general terms, and commits the organi::ation to y provide necessary resources, ~ 1 101 - I l
3
. . )
i Management Within the utility a specific individual should be named
.as Quality Assurance Officer (QAO) in charge of the QA work, with access to upper management and authority to recommend corrective action or work stoppage if necessary. Although the responsi-bility for quality assurance within the plant is shared by !all power plant personnel, upper management must stress the impor-tance of quality control and ultimately accept responsibility for the quality of ?.11 in-plant monitoring. Monitoring outside of the plant should be under independent management and each group should report all data directly tc the other.
The quality assurance plan should also specify the ' - ) individuals responsible for taking correct..ve actions. Some corrective measures, such as recalibration, may he cited, however ' many other less specific and more judgmental types of measures will not be documented. In most situations the professional knowledge and and technical judgment of staff 3rd outside consultants would dictate the appropriate corrective actions warranted by a situation. Management policy and staff dedication vAf* I and competen'ce would be vital to the effectiveness of the QA l i ! Program. l i
- -d
- 102 -
Objectives
'Next to management policy and staff dedication and competence the QA objectives are probably the most important part i of any project QA plan. They specify the precision, accuracy, completeness, reprecentativeness, and comparability necessary for each type of measurement (QAMS, 1980). -- j' ,
o Precision is the ability to obtain the same result i on repeated measurement. o Accuracy is the ability to obtain a true value, considering both precision and potential presence I of bias. o Completeness is defined as the percentage of , 1 sampling necessary to provide a valid result. Completeness is particularly important in evaluating field measurements which cannot be repeated. t i o Representativeness is. defined as the degree to l which measurements can be applied to the entire - i universe of possible meast;ements. j o Comparability efers to the types of measurements -j made and the Units in which they are expresned., l Standardization makes it possible to compare I results for like studies. 4
. ),
- 103 -
i Procedures ; I i The QA plar should describe specific procedures for each type of sampling including: sample custody, calibration , i procedures, sampling frequency, measurement procedures, minumum level of detection, and data reduction, validation, and reporting procedures. -- k' -; Sample custody involves procedures for identifying and maintaining a collected sample. so that the correct material is transferred and analyzed without any degradation or contamina- l tion. Calibration procedures and standards, and the required 3 frequency of calibration, should be determined 'and verified by
, daily checks on instrument performance. Calibration documenta- ' -
tion should also include instrument efficiency and, for 4 non-radiation measurements, the relationship between instrument readings and the measured parameter, such as flow rate. l Measurement procedures should be documented as Standard Operating Procedures. Measurement SOPS must be su*ficiently * , detailed so any two people taking the same type of measurement will follow the same procedures. SOPS must include checks to i show.that equipment is cperatin@ properly. i i J A central concern should be the minimum level of j detection necessary for particular radionuclides. The sample i 1 ,Q l i
- 104 -
J
size may be dependent on the minimum detection level as shown in Table 4. Data reduction, validation, and reporting procedures should be dwumented to include all data reduction schemes, and equations used in calculating results and in the statistical manipulation (e.g., averaging) of data. An example of thisj' is the mathematical method of correcting TLD readings for background. This section should also include the criteria for acceptance or rejection of outliers and wdys of verifying and checking computer calculations. It should cite data reporting methods including rounding, consolidations, omissions, and handling of negative valuer. All data should be retained in original form. Each result obtained should be included in , avetages, to avoid the biases introduced by excluding measure-ments that are less than the minimum detectable concentratien (Colle, et al., 1980.). Assecsment prccedures should be provided to reveal deterioration in measurement quality at an early stage before , , results are severoly af fected. The QA project shou 2d describe j the types of quality control (OC) measurements to be used and the criteria for determining it results are acceptablec in addition
; to internal quality control procedures, quality assessme3t ,
procedures should include. participation in intercompariscns, l measurecent of split samples, a'nd the level of degradation of 1 d 4 I 105 -
)
m -
- y. .. . . . _
1 I i Table-4 Coolant Detection Limits , Interference , Rep't Units (a) uCi/ml uCi/ml l Nuclide I H-3 2.-7 2.-7 C-14 5.-8 5.-8 Cr-51 5.-7 2.-5 t Mn-54 7.-8 5 -6
- Fe-55 1.-7 1.-7 .,
- -Fe-5g 1.-7 ~ 1.-7 i Co 3.-7 9.-6 Co-58 7.-8 S.-6 Co-60 1.-7 defined Ni-59 1.-7 1.-7 Mi-63 1.-8 1.-8 ;
Zn-65 2.-7 1.-5 Sr-89 2.v8 2.-8 ' St-90 1.-8 1.-8 Zr-95 1.-7 8.-6 Nb-94' 7.-8 4.-6 i Nb-95 7.-8 4.-6 i l Tc-99 2.-8 2.-8 .
- Ru-106 6.-7 3.-5 ;
) Ag-110M C.-8 3.-6 '
Sn-124 7.-8 3.-6 Sb-125 2.-7 7.-6 I-129 3.-0 3.-0 I-131 6.-8 2.-6 Cs-134 8.-8 5.-6 Cs-137 7.-3 4.-6 , )
- Ce-141 C.-8 2.-6 Ce-144 (gamma) 3.-? 9.-6
! Ce-144 (chem) 4.-8 4.-8 i U-235 or 238 1.-9 1.-9 Np-237 1.-9 1.-9 ;
Pu-238 1.-9
.-9 -
Pu-239, 240 1.-9 1.-9 Pu-241 8.-8 8.-8 Ac-241 1.-9 1.-9 Cm-244 1.-9 1.-9 l Am-243 (Np-239) 3.-7 9.-6 1
- Format example 2.-7=2x10-7 Source: Whalig, B.G., Walker, D.M., Gavi, M.R., Palma, J.M. 1987. :
i Appendix A, Characterization of Radionuclides in Nuclear l j Power Plant Effluents in ; r > Patrick, P. . and Palms, J. 1987. Light Water Reactors Muni- , toring and Management In-Plant and Environmental For Three l Mile Island, The Academy of Natural Sciences, Philadelphia,
- Pennsylvania. 'N' i I
- , , - , - - ---a
measurement quality that would require initiation of corrective ection. Quality Assurance (QA) Audits Provisions should be made for two kinds of QA audits, performance and system audits. The performance audit isasefof _ measuremants using ccrtified standard reference materials to obtain quantitative determination of the quality measurements. j The QA project plan should describe the measurement to be made, including the medium, the acceptance criteria, and the fraquency. A systems audit is an on-site qualitative examination and review of the measurement system by experts.. Audit reports should be sent to apper management. Here again, the effectiveness of the ,
. QA program would depend on the integrity and dedication of management.
f Freventive Maintenance The QA program should describe preventive maintenance . procedures and schedules. Auxiliary equipment is most likely tc i require periodic preventive maintenance. Incipient failures ! revealed by qua'.ity control review should be rectified by, correcti'.'e action rather than preventive maintenance. As the current generation of American nuclear power
)
plants ages, preventive maintenance is important to avoid y l
- 106 -
~-
[; ' ' , equipment breakdowns that lead to common mode failure and radiation accidents. For example, the sudden steam tube failu're in 1987 at Virginia Power's North Anna-1, PWR under cuts the long held assumption that equipment cracks develop slowly and can be detected by non continuous sampling before failure. (Inside N.R.C., November 23, 1987) Constant monitoring of cracks by physicallycheckingtheconditionofvalvesonin-linecontintEcus monitoring for steam tubes could pinpoint weaknesses and prevent common mode failure. Frequency of Reporting QA reports to management should be prepared on a quarterly, semiannual, or annual basis. These reports should ' , include an assessment of measurement precision, accuracy, and completeness; results of performance and system audits; corrective actions taken or required; and significant QA problems and recommended solutions. In addition, each project report should contain s QA section, relating the degree to which project results meet QA objectives. These reports should also be re-
- l l viewed by the NRC and by state and local authorities and made
~ #*' ~c" available for'public scrutiny and coniment. ~
l i I 4
- 107 -
t O Quality control (Qc) i Purpose Quality control is a component of quality assurance, but , it is result oriented. The main function of a quality control program is to verify that the associated quality acsura'hce program is operating properly. In a sense, the QA program provides policies, programs, and techniques that will give good results if properly effected; the quality control program assures that the quality assurance program is properly managed. 4 Quality control permits the recognition of difficulties ; - before they become serious. The American Society for Quality . , Control has defined quality control as "a continuing evaluation ! of the entire process with a view of having corrective action initiated as necessary" (ASQC, 1973). The nature of the corrective action cannot be defined in general terms an different types of measurements suggest different kinds of corrective action, and the correct course of action is determined by . involved professionals. ) An essential component of quality control programs is the establishment of criteria for performance and corrective i action. These criteria cover the desired precision and accuracy i of a specific measurement as well as acceptable variationc from l l the desired measuremen*.; for example, wnat percentage of d
-e
- 108 -
measurements beyond "two standard deviations" (from the mean) are I within an acceptable range for a specified measurement? ' Quality control has three functions: to assure managers and regulators of the validity of results; to detect deteriora-tion of measurement quality so that corrective action can be taken; and to satisfy others of the validity of results. hhe .l value of quality control for a radiation measurements laboratory is readily recognized. Procram Measurements and Procedures Because radiation monitoring results cannot be known in advance, a surrogate population of quality control measurements, normally distributed with a given mean and a given standard deviation, is established which can be replicated. Statistical examination of these measurements provides the necessary control information for the measurement system (Rosenstein, 1965). Control. charts are convenient tools for testing QC measurements. Shewhart (1931) established the criteria whereby a specific percentage of results in excess of two standard deviations from the mean (probability of about 5%) warn that a system may be getting out of control, and a specified percentage of results in excess of three standard deviations from the mean (probability
<0.5%) indicate that a system is out of control (these criteria are arbitrary).
- 109 -
a
. o I
Contribution of Quelity Assurance and Quality Control to Redundancy, Back-Up, and Safety , The Quality Assurance Program and associated quality i control methods provide some assurance that monitoring systems , ask the right questions and use the optimal monitoring. locations, frequencies, measurements, methods, calculations, and data ! handling procedures. Thequalityassurancephilosophysuggesksa ,f shift from the current emphasis of simply meeting regulatory standards, to that of using monitoring data to help limit, to the i environmental 1 greatest extent possible, the and human health ! impacts of commercial nuclear power plants. This concern for l l human and environmental health also emphasizes the advisability l of monitoring end effects. i e W e i
- 110 -
m _
e a
't CHAPTER 5: MONITORING END EFFECTS e
.g.
111 -
O O l The Need For End Effect Monitorinq ! l l In recommending the inclusion of end-effect monitoring, l the Advisory Board is well aware that the monitoring of humans and ecosystems to detect effects of exposures to ionizing radiation is not now included in nuclear power plant monitoring systems. Simila:ly, end-effect monitoring is not includedinfthe _ environmental monitoring programs for other facilities from which carcinogenic or toxic materials can escapes to the environment. The Board strongly suggests, however, that such monitoring should be routinely implemented, not only at nuclear power plants, but at other f acilities capable of releasing toxic materials to the environment either during routine operation or accident condi-tions. In making this recommendation, the Board does not suggest ' . that such monitoring be under the direct control of the operators of the facility, however the cost of end-effect monitoring, as other necessary compenents of the monitoring system, be borne by the firm which conducts the operation and thereby imposes unusual risk for its environs. The Board decided to include, this chapter on end-effect monitoring after the Fund's primary - consultants submitted their reports. Therefore this chapter is limited to an' overview of important consideratTCns'. i No matter how effective and complete a monitoring program is, the willingness of the local citizenry to accept nuclear plant advisories concerning public health impacts is
~
- 112 - (
5 a
limited because.of similar unwarranted assurances in the past.
been documented Clusters of childhood cancers have in areas around nuclear plants and these numbers are well beyond random i probability;6 yet the industry insists that radioactive nuclide releases from plants are insignificant and could not account for i these cancers. , Because of the above studies and because radiation cannot be detected by ordinary human senses, the public would like documented assurance that all avenues of transmission of ) radiation to their immediate environment have been adequately evaluated. The nuclear industry would be wise to recognize that, because potential health impacts of their operations are of such
- concern to the public, documented assurances must be provided. ,
- 6. Investigation of the Possible Increased Incidence of Cancer in West Cumbria. Report of the Independent Advisory Group.
Chairman Sir Douglas Black. HMSO, London, 1984. l Urgunart, J., Palmer, M. and Cutle r ,' J . , "Cancer in Cumbria: , The Windscale Connection," Lancet Vol. 1, 1984, pp. 217-218. . Heasman, M.A., I.W. Kemp, J.D. Urquhart and R. Black. Child-hood Leukemia in Northern Scotland. Lancet. Vol. I pg. 266. 1986. - Darby,'S.C. and Doll, R. Fallout Radiation Doses Near Dounrey and Childhood Leukemia. British Medical Journal, Vol. 294, j pg. 603-607. 1987. Norman E., Beral, V., Carpenter, L., Watson, A., Barton, C., Ryder, H., and Aston, D.C. "Childhood Leukomia in the West Berkshire and Baringstoke and North Hampshire District Health Authorities in Relation to Nuclear Establishments in the Vicinity," British Medical Journal, Vol. 294, pp. 597-602, , . 1987. -t 4
- 113 -
1 t e e l Types of Endpoints to be Monitored in
, Human and Animal Populations An exter sive public education program would be necessary to insure public agreement for an end-point monitoring program.
Two types of monitoring should be considered. The first would be measurements obtained from humans and animals to determine their burdens of radienuclides. Such measurements .would provideI a check on theorotical calculations derived from extensive monitoring of the environment and of food and water supplies. Detection of body burdens of some radionuclides, such as tritium, carbon-14, or plutonium-239 would be done by urine sampling and/or, possibly, teeth analysis and tissue sampling at autopsy. Other radionuclides, such as cesium isotopes, americium-241 and iodine-131, could be determined by whole-body ' - counting or measttrement of radioactivity in organs or body systems. Selecteil individuals or animals living close to the plant with know past history of residence and diet are the most likely to be exposed and should be monitored regularly. The seconc! type of end-effect monitoring that could be instituted includes continuing surveillance of vital statistics in communities near the plant. Adverse health effects from low doses of radiation include carcinomas resulting from irradiation at childhood or in utero and abnormal pregnancy outcomes such as stillbirthe or congenital abnormalities resulting from irradia-tion during pregnan:y. Detection of statistical changes in the - population of children or adolescents with carice r requires good
}
- 114 -
O . l l vital statistics or a cancer registry in the community. Detec-tion of abnormal pregnancy outcomes requires a network of information sources with cooperating doctors and hospitals in the l l nearby region. l Increases in car.cer rates near a nuclear facility would not be expected for many years af ter exposure to radiation, j' On the other hand, a significant change in the frequency of abnormal l pregnancy outr.omes might be obvious within one or two years. For both types of surveillance, an appropriate control population would have to be selected, preferably using established cancer or pregnancy outcome registries. l l Specific Preposals . , t , l Epidemiologic studies undertaken to investigate health l effects in relationship to radioactive discharges require that 1 l professional care be used in data collect'.n and analysis. Special attention must be given to the definition of the population under study, especially the effects of immigration and . l l emmigration over time. Anecdotal information may be provocative and occasionally insightful, but the public must understand that clustering'of a few cancer cases in a particular location or over a short period of time may occur by chance and does not l necessarily imply exposure to excessive amounts of radiation. If l nunprofessionals initiate an -investigation of potential . I environmental health hazards (usually at great personal sacrifice
- 115 -
l
O .0 and expense), they should make every effort to sec'tre profes-sional guidance. (Legator, et al. 1985) Cases of cancer should be collected within an area defined uy estimation of the probable distribution of airborne, waterborne, or foodborne radionuclides. Collection should begin as early as possible even before the nuclear plant is built, 2nd, . retrospectively, if data is available for the population under study. Coordination with the state cancer registry may help minimize costs. Data should be collected for at least three age groups: 0-4 years of age, 5-9 years of age, 10-14 years of age and, possibly, also 15-19 years of age. Care should be taken to define the populations at risk by age group; this should be relatively easy to do using birth and school records. The ' . percentage of population at risk that is selected for data collection may not need to be not be very large, but an adequate statistical power analysis should be used to determine how large. Table 5 shows the topic categories selected for a l- retrospective record review of adult and childhood cancers at Three Mile Island. i l
-t
- 116 -
O % Table 5 . Tcpic Categories For An Assessment Of The - Frequency And Pattern Of Cancer Events In The Area Around Three Mila Island (from Susser and Hatch, 1986)
- 1. Definition Of The Study Area Emissions Stress .
Other Environmental Exposures ~~ ( Available Demographic Data
- 2. Definition of the Time Period for the Study
- 3. Definition of Endpoints Radiation-related outcomes Stress-related outcomet Other Environmental-Induced Cancer
- 4. Choice of Data Sources Maximizing Ascertainment Record Review Procedures
- Data Items to be Abstracted S. Provisions For Patient Conflientiality
- 6. Data Abstraction Procedures Training the Head Abstractor Data on Cancer Mortality Correction of Residential Address Geographic Coordinates
- 7. Estimation of Exposure
- 8. Data Analysis Specification of a priori hypotheses .
Population Based Analyses Regression Analysis with Internal Comparisons Regression Analysis with External Comparisons Spatial Clustering Comparisons Not Based on Rates
- Pre-Post Accident Comparisons Time-Space Clustering Statistical Power to Detect Ef f er:ts
- 9. Criteria for Determining if a Follow-Up Case Control Study is Warranted s
~(
- 117 -
- 10. Approaches to Dealing with Limitations of the
- Proposed Study Study Area '
"Lead Time" Bias Population Estimates Multiple Sources of Radiation Exposure Residence As A Proxy for Radiation Dose In-Migrants Out-Migrants Movements Within The Study Area Place of Work i.
9 v 1 l l l l l l l s
-t
- 119 -
~
4 O To monitor pregnancy outcomes, recommendations include data collection of stillbirths and miscarriages during the first trimester occurring in the defined area. In both cases, special case-finding methods would have to be established. Accurate determination of the frequency of early miscarriages may not be possible, unless good cooperation with all local physicians can be developed. , '[ In addition, data should be collected concerning all cases of congenital abnormalities found among live births. Cases should be characterized by type and severity, and rates per number of live births. Again cooperation with local hospithis and physicians will be necessary to ensure completeness of records. The statistics generated can be compared to data - , available from the Centers for Disease Control in Atlanta, which has been collecting such data from many centers in the U.S. for years. For all these endpoints, childhood cancers, stillbirths . and congenital abnormalities, an ongoing investigation should be - maintained. Efforts must also be made to determine whether other environmental or occtya91onal factors may be present which could influence cancer or pregnancy outcomes.
- N,
- 119 -
Genetic Effects on Plants It is generally not possible to defect an increased rate of genetic or hereditary effects due to low level radiation exposure in human or animal populations. Except in very large populations of species which have similar sensitivity and are exposed uniformly to relatively large doses of radiation. f _ There is, however, one well-studied plant species, the 4 spiderwort plant (Tradescantia), in which genetic effects at low doses are relatively easy to detect within three weeks of exposure. A mutation causes a change in the color of individual stamen hair cells, of selected strains of heterozygous blue-pink spiderworts. These colon changes can be easily seen under a . microscope. Use of spiderwort mutations to monitor airborne or deposited radionuclides emitted from nuclear power plants has already been done effectively in Japan by Ichikawa (1981). l The Board recommends that Tradescantia monitoring be done by scientific groups in universities or high schools near the power plants. Spiderworts of the proper strain could be grown in locations around the power facility and harvested regularly ~for microscopic examination. Spiderworts growing upwind or in unexposed areas could be used for controls. , Monitoring groups must have the technical expertise to grow and examine the Tradescantia under reasonably controlled situa-tions. The mutation rate in these plant cells would reflect the }
- 120 -
0 O. 2 integrated sum of recent and or long past exposures to radiati.on and or to other mutagens in the environment. , In summary, a number of end-effect evaluation techniques . used on plant and animal populations living around nuclear facilities, could be valuable in determining mutations, childhood cancers, or abnormal pregnancies. The absence of abno r',;nal findings could be confirmatory evidence of the reliability of in-plant and environmental monitoring of radionuclide releases. Conversely, the presence of abnormal findings would be cause for review of plant operations and monitoring data.
-t
- 121 -
[ C 6 4
.. i.' .
CHAPTER 6: ROUTINE AND EMERGENCY DATA COMMUNICATIONS e N
- 122 -
1 Communications concerns 7 - l ! During an emergency, the recommended comprehensive l monitoring program would provide information in an expedient I manner to the plant manager; local, state and federal agencies; and eventually summary information to the public. During routine operations, the recommended program would provide informah' ion that would permit these authorities to realistically assess any associated risks, and inform the public. l l Current Communications System l Current regulations as well as the actions of private i groups have defined the current system of data communication. ' . Data Communications During Routine Operations 1 l The Public Record -- By law, all documents related to I regulation of a nuclear power plant by the federal government must become part of the pinblic record or that body of information - to which the public has legal access. A formal docket and docket
. number are assigned to each nuclear plant file, ar.d all i l
- 7. (This section is based on Appendix I, "Communications" by i
David M. Walker in Light Water Reactors Monitoring and j Management In-Plant and Environmental For Three Mile Island,
- Academy of Natural Sciences, Philadelphia, 1987 and a ,
forthcoming article, "How The News Media Reported On Three l Mile Island and Chernobyl" by David M. Rubin. Brief -l quotation from this article is by permission of the author j and the editors of the Journal of Communication in which the '-
?
full text of the article appears in the Fall 1987 issue.) 1
- 123 - j
appropriate material is placed onto the docket by the Nuclear Regulatory Commission. (Walker, 1987) - The semi-annual effluent release reports for operating plants are available in NRC public reading rooms. These reports include summary information on all routine releases of liquid, gaseous, and solid radioactive effluents from a. site. Utilikies are required to submit a report to the NRC within 30 days if a Limiting Condition for operation has been ex:eeded. Other channels -- occasionally, a utility will provide < reports directly to the state environmental agency, and issue press releases to the news media which summarize the data from the perspective of the utility. Special interest groups receive , radiological effluent and environmental information either directly from the utilities, or from public information. These agencies also issue reports, statements, and press releases summarizing data according to their perspective. In the Three Mile Island area, the TMI Alert organization issues a newsletter entitled, Island Updates, while GPU issues the GPU Nuclear - Newsline. The NRC also issues a newsletter on the progress and monitoring of the clUt:Eup activities of TMI-2. 4 At the present time, the news media provide the primary link with the public for the dissemination of information concerning radiological consequences resulting from routine plant Because the news media generally do not have the s operations. -t ] - 124 -
technical resources to interpret effluent and environmental reports, the information conveyed to the public usually consists of edited versions of utility, regulatory agency, or special interest group press releases. As such, the information presented to the public carries the perspective of the special ; interest group which provided the news release (Walker 1987). Direct communication of routine effluent and environmen-tal monitoring results to the public by the utility is not widely attempted in the U.S. Attempts by the utilities to include material in routine mailings, such as utility bills, have met with strong opposition and lawsuits by anti-nuclear groups, who contend that the utility presents selected "good news" which unfairly supports their own businese interests (Walker, 1987). + , A few U.S. utilities use a positive information approach ; like that employed in France to convey radiological information in a credible manner to the public (Vendryes, 1985). Dr. Vendryes, a former Chairman of the Atomic Energy Commission of France, explains that at each nuclear plant site in France an - information commission is created at public request comprising elected officials and representatives of local unions and associations (Wal'<er, 1987). In September 1987, France's High
+ Council For Nuclear Safety and Information (CSSIN) as a result of ,
the controversy over the Chernobyl reporting recommended that the i i 1 government's electionic nuclear information magazine, MAGNUC 4 l publish information from competent unofficial sources in addition
- _g:
- 125 - l
. o to government communiques. The council also recommended that an i ad hoc expert group be asked to establish a system of objective criteria for classifying nuclear events for their relative importance and that MAGNUC's articles be updated more of ten and
- written in two versions, one for the general public and the other for a more knowledgeable audiance. The group further proposed that, in an emergency, dailyortwicedailypress, conferences}'be .
held at pre-set times to establish more direct permanent contact with the press (Nucleonics Week, 1987) Data Communications During Emergencies The TMI-2 accident demonstrated the need for a carefully conceived, pre-tested communications plan. That communication , breakdowns contributed to the severe impact of the TMI-2 accident is well-documented. Communication problems during the accident occurred at all levels, from information exchanges among plant workers, to information provided by the NRC to the President of the United States (Walker, 1987). The TMI-2 accident also resulted in media distrust of government and utility supplied information (Rubin, 1987). The media response to nuclear accidents continues to be ad hoc. Most reporters assigned to cover the nuclear power industry have little understanding of risk assessment, the radiation ! vocabulary, or nuclear operations (Wenger, 1987). s
*k
! - 126 -
Rubin's- 1987 review of information flow and media coverage of the Windscale, TMI-2, and Chernobyl accidents show's that "f rom a journalist's perspective, a serious nuclear power plant accident is wholly unlike other disasters such as hurri-canes, floods, or earthquakes. Journalists are naturally suspicious of anything nuclear because of the secrecy surrounding the t,echnology and the lack of candor all governments have shown , in the face ot .ccidents" (Rubin, 1987). The historic lack of candor about radiation releases has forced journalists to seek other sources of information. Since there are "few recognizably neutral" sources for reporters to consult.", The end result has been inaccurate reporting and an unwillingness of reporters to assess the "long-range ef fects of their work" (Rubin, 1987). This has further fueled the untility/ press hostilities and helped to create a distrustful public. The prevailing media view of emergency communication is i i to "print the news and raise hell" combined with faith that the audience can sort out conflicting information and act rationally. (Rubin, 1987). Despite the prevailing suspicion between those who supply' radiation data and those who repcrt it, the public c requires specific monitoring information. l l l During an emergency at a nuclear power plant, the public is served' by two types of communications. First, information
.}
- 127 -
i describir.g present and projected radiological conditions and consequences must be made available to those charged with pro-tacting the public from immediate or subsequent harm. ' Secondly, members of the public must be alerted that an emergency exists, must be given unambiguous instructions as ';o protective actions, and must be provided with timely information concerning the nature and significance of any radiological releases ( Walk'e r , 1987). On a technical level, several types of information must be available to emergency response teams to permit them to assess radiological conditions, and project consequences. For the purposes of this discussion, the tecnnical information needs are grouped into four categories. Table 6 (Walker, 1987). , , I l , i I I i i
-b l
- 128 -
- ,--,w-,.. , - - - _ , _ . - - - _ , ,, _ _ , , _ _ , , _ _ , _ , _ _ _ _
j
} .
Table 6
- Radiological Information Needs For Response to Emergencies at a.
Nuclear Plant Site (from Walker, 1987)
- 1. On-site Data
- condition and state of the reactor
- radionuclide identification and concentration data for any liquid or gaseous volume which may.,be released from the site .. !
- readings of process, area, and effluent radiation monitors
- actual and/or possible release rates of radioactive materials into the environment
- meteorological information
- 2. Off-site Data
- dose rates, cumulative dose, iodine concentrations, and radionuclide information for different physical forms of effluents at known ground level locations
- dose rates, iodine concentrations, and radionuclide characterization of the plume at known locations
- radionuclide concentrations in foodstuffs, water, '
milk, and animal foods (long term)
- 3. Calculated and Estimated Results
- present and projected local weather behavior
- present and projected dose levels and radionuclide deposition at off-site locations
- 4. Pre-determined Information and Criteria
- radiological source term and release rate estimates *
- geographical distribution of the population ao a function of time of day
;
- criteria for implementing prescribed protective actions.
l s
..{
4
- 129 -
O 4 l Current regulations establish the minimuni necessary l requirements related to the generation and communication of 1 monitoring information following a nuclear accident. NUREGS-0654, 0737, and 0654/ FEMA (Federal Emergency Management Agency) provide guidance for emergency communication facilities, necessary equipment and staff, the timing of notification, and l l press facilities. Because there 's no current standard ,for procedures and facilities, utilities have developed a variety of options for communicating emergency data. Provisions for acquiring and processing emergency l radiological data vary greatly. Verbal communication, handwritten records, and "whiteboard" data displays are more frequently used than are the electronic digital transfer, recording, and display , techniques commonly used in other industries (Walker, 1987). Several utilities interpreted the regulatory guidance l advanced by the NRC during the immediate post TMI-2 period as advocating electronic data gathering, transmission, storage, and display. Some utilities developed and installed state- of-the- . art electronic information systems, only to find that they are out-of-step -with thWMC and other government agencies, which continue to use verbal communications and paper recorde systems (Walker, 1987), s
.*(
- 130 -
- f. ,
MET PLANT OFFSITE OTNER TOWER MONITORS MONITORS MOBILE LAB MOBlLE als MONITORS C Ol 0 - -
'l* -{-] _._
-a -
E'e' 'e. INPUT MIA sUtmTIM
+44++
l l
/I I 1
.:" " / mim \
cm ,,,,, h SYSTEM CONSOLE
, a ~ . .
. = =;
DCPP TSC CC CENTRAL COMPUTER 120 MBYTE DISK
** 8mS 5cas ecam ewamroe osumnews h ;\ /r ' ' '\\ b A p A /m\
l rio ca coma. n eGDer conmwi coomu su m m7 Roce cramms acroT o mc or CDCD FAduTY RESPoNsf DOGDCY l can soms 1 Figure 12: Diablo Canyon Power Plant Emergency Assessment , and Response System From Walker,1983
The NRC response center is still a maze of telephones and white boards, and the NRC and FEMA computers for emergency data have incompatible software and cannot communicate. Communication of data between the NRC Response Center in Bethesda and other federal emergency response agencies is still done verbally. This information distribution process would be well-servedbyanelectronicdatagatheringsyst,emwith"bullIetin . board" display to which any agency could post current information or receive status updates (Walker, 1987). The Nuclear Data Link (as described in NUREG/CR-1451) was to provide a direct electronic link to the NRC for plant paramoters from each site; but additional requirements, budgetary constraints, and strong opposition by utilities hindered * , development. A revival of the Nuclear Data Link in simplified c form is now being pursued by the Incident Response Branch of the
- NRC, and has been relabeled as the Emergency Response Data System (Walker, 1987).
l An example of a computer based "NUREG-0654" system was - installed at the Diablo Canyon Power Plant in 1980 and is shown i in Figure 12 (Walker, 1983). A central computer automatically collects and archives data from monitors at on-site and off-site locations and simultaneously displays the information at color graphics workstations in the EOF, TSC, control room, Corp 6 rate Incident Response Center (200 miles away), and California State l Office of Emergency Services (250 miles away). Data can be '
--t 1
- 131 -
l
O O , independently reviewed from each work sta'lon c and dose l projections can be transfered onto selected maps for one or fo.r multiple stations (Walker, 1987). The Diablo Canyon system has been in routine full-time i operation for five years and plans are in progress to upgrade system software and hardware. In 1982, theDiabloCanyonsys)em l was operated from a workstation located at the NRC Bethesda ! offices in an NRC demonstration. However, most utilities : r strongly resist direct on-line links to plant data by the NRC. The nuclear power industry concern is the possible misuse of such ; data by remote NRC participants during an actual emergency situation (Walker, 1987). l Problems With Current Data Communications record ~ The public for nuclear plant operations is !. difficult to access due to: incomplete docket files, delayed data ! filing; and limited availability. The NRC public reading rooms 2 for nuclear plant docket materials are used extensively by the
; special interest groups, to a lesser extent by the media, and j 1
[ i very infrequently by individuals (Walker, 1987). N This passive means of data communication does not
=
l provide the public with timely information and is not suffi-a ciently effective to serve as the primary means of communicating monitoring information. Information and reports required as part I I of the regulatory process also do not provide a complete and or timely record of effluent monitoring (Walker, 1987). 4 J - 132 - i
o . Emergency Operations Covernments and utilities historically have tried to present an optimistic picture of a nuclear emergency. This attitude has increased the p'ablic and press suspicion of utility generated communications. In addition, the press has demonstrated that it may not have the technical experb'ise necessary to permit accurate reporting of nuclear industry data, t Information Priority and Digital Equipment Many emergency response plans require the collection and communication of large amounts of radiological data, often without regard to the relative importance of each item of ' . information. A distinction can be made between essential data and non-essential data. Such distinctions must be made in advance of an emergecy situation, since emergency conditions are less than optimal. It is important to understand that the method of communication affects the. dependability of data (Walker, 1987). The tempo during an emergency may well range frem intense - activity to long periods of routine data collection and review. During intense activity, verbal request / response exchanges are particularly ineffective and disruptive. Electronic access to the same data is much more efficient and reliable. 4
-t
- 133 -
V Dose Projection Determination of a credible radienuclide-specific release rate for dose projections is difficult to accomplish in a timely manner during accident conditions. Because dose projections are needed promptly at the onset of an accident, it is important that a carefully pre-planned default sequence {' be available to assure that release rate data is available for dosage calculations. This default sequence might include measured radionuclide concentrations, fixed monitor readings, and calculated source-term inventory based on reactor power history. Because radionuclide-specific information is essential to dose calculations, primary and backup methods should be l l instituted for promptly obtaining and communicating postaccident ,, t 1
- l. gamma spectrometry measurements (Walker, 1987).
l 1 Off-Site Data l l Off-site radiological measurements at ground based, l locations are essential to confirm the nature and magnitude of . radiation related to accidental releases. Off-site readings are assigned a greater %.c@ibility by both the public and response team members than are calculated or projected values. Currently field measurement values are transmitted by radio and subject to poor transmission quality and operator errors (Walker, 1987).
-t
- 134 - l l
i
Radiation tracking of. radioactivity in gaseous plumes 4 can be accomplished most effectively by airborne monitoring systems. The complex three-dimensional geometry of radioactivity in a plume cannot be effectively measured by the limited sensitivity and radiation response capability of ground based detector arrays.
'l-Proposed Improvements In Data Communication In-Plant Improvements The electronic technology available for recording, transmitting, storing, and retrieving data provides fast, + ,
accurate, and highly dependable data handling. If large amounts of data are visually displayed to the participants, the data should be prioritized, by the use of color or other graphics techniques, to assure that important deviations will be noted promptly. By contrast, manual or verbal data processing under ! emergency conditions introduces variable delays with a high - 1
- probability of error (Walker, 1987).
l A new computer system will centralize TMI in plant effluent data and provide management with ready access to reliable monitoring data. In addition, if the off-site, remote, real-time on-line data is displayed in the control room and - I electronically linked to the . meteorological data base and the j 1 - 135 -
in-plant. effluent data, in the manner of the Diablo Canyon System, then the operators will have a powerful tool to analyse the implications of releases. There should also=be a real-time on-line link to the county emergency manager's office. Public Access Improvements l' The following suggestions for the improvement of timely ; public access to monitoring data are not exhaustive. Additional discussion may result in more suitable approaches. 1 Citizen Involvement -- Local citizen representatives , should be involved in planning a comprehensive radiological environmental monitoring program. The final adopted program ., l should be offactively and comprehensively communicated to the i public (Walker, 1987). i Wherever possible, local citizens (farmers, gardners) should be involved in the environmental sampling and measurement l l program (Walker, 1987). A recognized review group of local . i citizens, some with scientific credentials, should be designated l ! to receive all routine reports, and should be kept informed of l . the progress of the environmental programs (Walker, 1987). . i 1 ! Public Monitors -- Dynamic radiation measurement devices ,
! should be placed near population centers where trained citizens i
4 can obtain immediate information about ambient radiation levels ,
.g J
- 136 -
1 j
.i_.____..___. _ _ _ _ _ _ _ _ _ . _ _ _ - _ , m ____ . . . _ . . . _ _ , , . , . _ _ _ _ _ _ . _ _ _ , _ _ _ _ . _ . _ , _ , _ _ , . _ _ . . - _ _ _ .. __ , _ , _ , . _ . .
- 4 .
(Walker, 1987). Wherever appropriate and feasible, local qualified citizens should take independent environmental
- radiation measurements. This could be done by local high school
! or college science teachers and students (Walker, 1987). 1 Routine and Emergency Data I i Improvements must be made in the transmission of i environmental radiation data to the utilities, governments,'and the public. Data from instruments, such as the PICS (pressurized ion chambers) which dynamically measure ambient radiation, should 3 be transmitted electronically using dedicated phone lines, FCC i licensed radio frequencies, and satellite based systems (Walker, 1987). ' . p i The NRC should improve the manner in which the docket process is structured and maintained, and should provide current l i summaries of effluent and environmental monitoring program j results. l Utilities should be encouraged to provide up-to-date
- summaries of effluent and environmental monitoring program results in an understandable format for the general public l
I (Walker, 1987). l 4 ) l -D l l - 137 -
. s Independent Monitoring Groups The public distrust of the nuclear power industry creates a communication gap that could be filled by independent monitoring. Currently at TMI, there are over two thousand identified emergency workers who have access to an equal number of calibrated high resolution survey meters provided by kEMA (Pennsylvania Emergency Management Agency) and the utility.
These are only issued during an emergency; however, PEMA could release these instruments for use in a coordinated, calibrated, independent off-site program. Such a program would require training and coordination to assure the most ef fective use and maintenance of equipment. , - Patrick and Palms (1987) recommend that an independent academic group conduct the bicaccumulator sampling program. Similarily, the TMI Public Health Fund Advisory Board recommends independent monitoring of end-effects. Funding could be made available through state and local governments, foundations, taxes on the utility or outright grants from the utility. Should the - utility endow such an independent "no strings attached" effort, it is likely that credibility of nuclear industry communications j would improve. l The charter of the independent bicaccumulator and sutvey meter monitoring group would provide for monthly or quarterly . summaries of data to be published in the local newspapers using *!
.-q
- 138 -
o e t i i understandable language. Media representatives might also i benefit from these independent summaries as .they would provid'e i familiarity with the vocabulary of radiation measurement and dose r j assessment. Independent monitoring and reporting will serve as a ! j check on communication releases from the utilities and increase ; I j public confidence in industry communications. l , Media Considerations , Rubin believes "governments and private utilities can shape public information policies so the needs of the news media ! i
- are met while the worst excesses are avoided." When dealing with l 3 r the media, government regulatory agencies and, nuclear power l
utilities should determine in advance of accident conditions: *. 4 e ! l ' (1) How radiological information will be made public in [ i ! a timely and understandable fashion. , (2) Guidelines on the centralliation of information and I at what phase of an accident information should be released. - 1 l l ' *# M - (3) Who will serve as press liaison, and the role of I management in briefings. i l . (4) A policy on "gag orders." l - l } s j . -t l i
- 139 -
1
. . 1 (5) Policy concerning how close journalists will be permitted to get to a damaged reactor. .
i During normal plant operations, utilities and regulatory agencies should develop policy concerning: , (1) The establishment of press seminars on radiaq' ion monitoring, risk reporting, and nuclear engineering. In addition, policy should stress an ongoing update process for reporters who cover nuclear affairs. , (2) Whether or not local journalists will be given l I pride of place over national and foreign journalists; and whether specific press organizations deserve information first and in , [ what detail. ; i (3) "mergency information plans for each power plant. G
-vv d.' .
t
, .z
- 140 -
e O e 9 4 h 5
-e e
CONCLUSION o 9 4 k
-t 141 -
o e A radiation monitoring system for commercial nuclear power plants must provide data for plant operation from: in-plant measurements; environmental samples; and studies of human, plane, l and animal end-effects. These three data sets must provide
- aufficient back-up and overlap to monitor all pathways of exposure even in severe accident conditions. They must also 4
provide the public with assurance of compliance with regulati,6ns y established to protect public healt? - assessment of risks from t normal operations, and timely we . .g in the event of an accident. current radiation monitoring systems at me e 'i . s . I commercial nuclear power plants, and TMI-l in particular, meet NRC regulations at optimal operation; but even with optimal *, operation and NRC compliance, the system has many gaps in i coverage. Implementation of the three-part monitoring system I recommended in this report should close these gaps. However, overlapping measurements alone will not improve reactor j performance. Both regulators and utility management must change 1 8 their view of the function of monitoring from the verification - I of regulated limits to that of a centrali:ed data gathering and i reporting safety net which can be analyzed to: continually reduce j emissions, provide advance warning of operational weaknesses, and } provide the public with accurate timely data monitoring reports. l
-t l - 142 -
g M P e 1 REFER :4CES CITED 3 l l l
=
I s
..k 1
- 143 -
1 I l 1 t
w American Society for Quality control, Statistical Control Committee. 1973. Glossary and Table for Statistical Quality
- Control. ASQC, Milwaukee, Wisconsin.
Bell, M.J. 1973. ORIGEN, the ORNC Isotope Generation and Depletion Code. Oak Ridge National Laboratory Report, ORNC-4628, (Revised by S.M. Stottler Corporation, June, 1977.) Beyea, J. 1984. A Review of Dose Assessments and Recommendations for Future Research at Three Mile Island. Three Mile Island Public Health Fund, Philadelphia. BIER I. 1972. AdvisoryCommiteeontheBiologicalEffectsfof Ionizing Radiation (BEIR Report). National Academy ~of Sciences - National Research Council, U.S. Gov't. Printing Office, Washington, D.C. Christianson, D.R. 1984. Empirically Determined Airborne Particulate Plateout in Radiation Monitor Sample Lines. Trans. Am. Nuc. Soc., 47, 389. Colle, R., H.R. Abee, L.K. Cohen, D.Ed., E.H. Eisenhower, A.R. Jarvis, I.M. Fisenne, M. Jackson, R.H. Johnson, Jr., D. Olson, and J. Peel. 1980. Reporting of Environmental Radiation Measdrements Data, in Upgrading Environmental Radiation Data. A Report of an Ad Hoc Health Physics Society Committee (J.E. Watson, Chairman). U.S. Environmental + . Protection Agency Report EPA 520/1-80-012). " Eisenbud, M. 1973. Fnvironment Radioactivity, Second Edition, Orlando: Academic Press. England, R.W. 1987 Estimates or Environmental Accumulations of Radioactivity Resulting From Routine Operation of New England Nuclear Power Power Plants (1973-84) A Report of the Nuclear Emissions Research Project Wittemore School of Business and Economics, University of New Hampshire, Durham. Franke, B. 1987. Sensitivity Analysis of Monitoring - Programs For TMI: Report to the TMI Public Health Fund. Takoma Park: Institute for Energy and Environmental Research. Ichikawa, S. 1981. "In situ Monitoring wit' Tradescantia Around Nuclear Power Plants." Env. Health Pr:spectives, Vol. 37, 1981 pp. 145-164. Inside, N.R.C. "North Anna Rupture Spurs U.S. Interest in EDF's New Leak' Detector" Vol. 9, No. 24, P. 13, November 23, 1987. McGraw - Hill Inc. NY. NY. Kabat, M.J. 1982. Deposition of Airborne Radioiodine Species on Surfaces of Metal and Plastics. Proceedings of
- the 17th DOE Nuclear Air Cleaning Conference, 17, 285. 4
- 144 -
, e Kahn, B. 1980. Composition and Measurement of Radionuclides in Liquid Effluent from Nuclear Power Stations. Effluents and Environmental Radiation Surveillance, ASTM STP 698, '
63-74. Kirk, W. June 5, 1987, "Termination of Remote Telemetering of Ambient Gamma Radiation from EPA's RSS 1011 "Sentu" System at TMI." U.S. Environmental Protection Agency, Middletown, l Pennsylvania. Legator, M.S. Harper, B.L, and Scott, M.J. 1985. The Health Dectective Handbook: A Guide to the Investigation of Environmental Health Hazards by Nonprofessionals. Balti@ ore Maryland: Johns Hopkins University Press. - l Maeck, W.J., et al. An Assessment of Off-Site, Real-Time l Dose Measurement Systems for Emergency Situations, Report ( NUREG/CR-2644, U.S. Nuclear Regulatory Commission, ' Washington, DC 20555, 1982. Miller, D. 1983. "Safety Measures at Commercial Nuclear Power Plants," Public Forum on Nuclear Power, March 28, 1983, l Middletown, Pennsylvania. l Nucleonics Week, "Recommendations on Improving the Channels of French Nuclear Information", Vol. 28 No. 45 Nov. 5, 1987
- p. 8. McGraw-Hill, NY. NY. ,
Nurrenbern, J.D., and R.R. Roselius. 1985. Sample Line
- Plate-Out Measurements on the Unit Vent Wide Range Gas Monitor at the Calloway Nuclear Plant. Presented at the 30th Annual Meeting of the Health Physics Society, Chicago, May 26-31, 1985.
Paciga, J. 1985. Telephone Communcation with Dr. John Paciga, plant staff, Point Lepreau Nuclear Station, Point Lepreau, New Brunswick, Canada, November, 1985. Patrick, R- 1987. Environmental Monitoring, Appendix B in . Light Water Reactors Monitoring and Management In-Plant and Environmental For Three Mile Island, The Academy of Natural Sciences, Philadelphia, Pennsylvania. f Patrick, R. and Palms, J. 1987. Light Water Reactors Monitoring and Management In-Plant and Environmental For Three Mile Island, The Academy of Natural Sciences, Philadelphia, Pennsylvania. Patrick, R. and Palms, J. 1987. Section A, Chapter 1: Objective of an Environmental Monitoring Program, in Light Water Reactors Monitoring and Management In-Plant and Environmental For Three Mile Island, The Academy of Natural Sciences, Philadelphia, Pennsylvania. s
...t
- 145 -
,7 _ . - .
Chapter A2: Fundamental Questions and Concerns in Designing a Comprehensive, Effective Monitoring Program. , Chapter A3.2: Environmental Monitoring. Section B: In-Plant Monitoring of Possible Radioactive Effluents. Chapter B4: In-Plant Effluent Monitoring. Preston, D.L. and Pierce, D.A. "The Effect of Changes in Dosimetry on Cancer Mortality Risk Estimates in the Ato,mic Bomb Survivors" Radiation Effects Research Foundation, TR,j 9-87, 1987. Quality Assurance Management Staff, 1980. Interim Guidelines and Specifications For Preparing Quality Assurance Project Plans, U.S. Environmental Protection Agency Staff Report. QAMS-005/80. Rosenstein, M., and A.S. Goldin. 1965. Statistical Technics for Quality Control of Environmental Radio-assay, Health Laboratory Science 2, 93-102. Susser, M., and Hatch M. 1986, Protocol For Phase II of the Three Mile Island Cancer Study, Preliminary Assessment of the Frequency and Pattern of Cancer Events in the Area Around * , Three Mile Island, The Gentrude H. Serqieusky Center Columbia University for the Three Mile Island Public as filed of record with the United States District Court, Harrisburg, Pa. September, 1986. Shewart, M.A. 1931. Economic Control of the Quality of Manufactured Products. Van Nostrand Reinhold Co., Princeton, New Jersey. Till, J. 1987. Appendix C, Mathematical Models as the basis For Monitoring Program Design in Light Water Reactors Monitoring and Management In-Plant and Enviornmental For - Three Mile Island, The Academy of Natural Sciences, ! Philadelphia, Pennsylvania. 1 U.S. Nuclear Regulatory Commission. 1979. TMI-2 Lessons Learned Task Force Status Report and Short-term Recommendations. NUREG-0578, July, 1979. U.S. Nuclear Regulatory Commission. 1982. Correction for l Sample Conditions for Air and Gas Monitoring. IE Information Notice No. 82-49, December, 1982. United States Environmental Protection. 1979. Quality Assurance Poligy Statement, Environmental Protection Agency, w
._,k
- 146 -
L
9 , ,. Vendryes, . 1986. Observations on the Nuclear Programs of France and the United States, paper prepared for the Atomic Industrial Forum. , Walker, D., 1983. Implementation of an Emergency Assessment System, presented at the IEEE 1982 Nuclear Power Systems Symposium, Washington, D.C., October, 1982. Walker, D. 1987. Appendix I, Communications in Light Water Reactors Monitoring and Management In-Plant and Environmental For Three Mile Island, The Academy of Natural Sciences, Philadelphia, Pennsylvania. , t Wenger, D. 1987. Direct communication with- Jonathan Bercjer Whalig, B.G., Walker, D.M., Gavi, M.R., Palms, J.M. 1987. Appendix A, Characterization of Radionuclides in Nuclear Power Plant Effluents. { Whalig, B.G., Walker, D.M., Gavi, M.R., Palms, J.M. 1987. f Appendix E, In-Plant Monitoring Systems for Radioactive l Effluents from Nuclear Power Plants in Light Water Reactors 1 Monitoring and Management In-Plant and Environmental For Three Mile Island, The Academy of Natural Sciences, Philadelphia, Pennsylvania. . l 1 tmiS053
'l
.-y
{ l
- 147 -
;}}